diff --git "a/log/log-train-2023-02-08-23-42-53-1" "b/log/log-train-2023-02-08-23-42-53-1" new file mode 100644--- /dev/null +++ "b/log/log-train-2023-02-08-23-42-53-1" @@ -0,0 +1,2920 @@ +2023-02-08 23:42:53,784 INFO [train.py:973] (1/4) Training started +2023-02-08 23:42:53,784 INFO [train.py:983] (1/4) Device: cuda:1 +2023-02-08 23:42:53,850 INFO [train.py:992] (1/4) {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 50, 'reset_interval': 200, 'valid_interval': 3000, 'feature_dim': 80, 'subsampling_factor': 4, 'warm_step': 2000, 'env_info': {'k2-version': '1.23.3', 'k2-build-type': 'Debug', 'k2-with-cuda': True, 'k2-git-sha1': '3b81ac9686aee539d447bb2085b2cdfc131c7c91', 'k2-git-date': 'Thu Jan 26 20:40:25 2023', 'lhotse-version': '1.9.0.dev+git.97bf4b0.dirty', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'surt', 'icefall-git-sha1': 'b3d0d34-dirty', 'icefall-git-date': 'Sat Feb 4 14:53:48 2023', 'icefall-path': '/exp/draj/mini_scale_2022/icefall', 'k2-path': '/exp/draj/mini_scale_2022/k2/k2/python/k2/__init__.py', 'lhotse-path': '/exp/draj/mini_scale_2022/lhotse/lhotse/__init__.py', 'hostname': 'r8n07', 'IP address': '10.1.8.7'}, 'world_size': 4, 'master_port': 12354, 'tensorboard': True, 'num_epochs': 30, 'start_epoch': 28, 'start_batch': 0, 'exp_dir': PosixPath('pruned_transducer_stateless7_streaming/exp/v1'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'base_lr': 0.05, 'lr_batches': 5000, 'lr_epochs': 3.5, 'context_size': 2, 'prune_range': 5, 'lm_scale': 0.25, 'am_scale': 0.0, 'simple_loss_scale': 0.5, 'seed': 42, 'print_diagnostics': False, 'inf_check': False, 'save_every_n': 2000, 'keep_last_k': 10, 'average_period': 200, 'use_fp16': True, 'num_encoder_layers': '2,2,2,2,2', 'feedforward_dims': '768,768,768,768,768', 'nhead': '8,8,8,8,8', 'encoder_dims': '256,256,256,256,256', 'attention_dims': '192,192,192,192,192', 'encoder_unmasked_dims': '192,192,192,192,192', 'zipformer_downsampling_factors': '1,2,4,8,2', 'cnn_module_kernels': '31,31,31,31,31', 'decoder_dim': 512, 'joiner_dim': 512, 'short_chunk_size': 50, 'num_left_chunks': 4, 'decode_chunk_len': 32, 'full_libri': True, 'manifest_dir': PosixPath('data/manifests'), 'max_duration': 500, 'bucketing_sampler': True, 'num_buckets': 30, 'concatenate_cuts': False, 'duration_factor': 1.0, 'gap': 1.0, 'on_the_fly_feats': False, 'shuffle': True, 'drop_last': True, 'return_cuts': True, 'num_workers': 2, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'input_strategy': 'PrecomputedFeatures', 'blank_id': 0, 'vocab_size': 500} +2023-02-08 23:42:53,850 INFO [train.py:994] (1/4) About to create model +2023-02-08 23:42:54,155 INFO [zipformer.py:402] (1/4) At encoder stack 4, which has downsampling_factor=2, we will combine the outputs of layers 1 and 3, with downsampling_factors=2 and 8. +2023-02-08 23:42:54,167 INFO [train.py:998] (1/4) Number of model parameters: 20697573 +2023-02-08 23:42:54,168 INFO [checkpoint.py:112] (1/4) Loading checkpoint from pruned_transducer_stateless7_streaming/exp/v1/epoch-27.pt +2023-02-08 23:43:03,616 INFO [train.py:1013] (1/4) Using DDP +2023-02-08 23:43:03,872 INFO [train.py:1030] (1/4) Loading optimizer state dict +2023-02-08 23:43:04,090 INFO [train.py:1038] (1/4) Loading scheduler state dict +2023-02-08 23:43:04,091 INFO [asr_datamodule.py:420] (1/4) About to get the shuffled train-clean-100, train-clean-360 and train-other-500 cuts +2023-02-08 23:43:04,266 INFO [asr_datamodule.py:224] (1/4) Enable MUSAN +2023-02-08 23:43:04,267 INFO [asr_datamodule.py:225] (1/4) About to get Musan cuts +2023-02-08 23:43:05,805 INFO [asr_datamodule.py:249] (1/4) Enable SpecAugment +2023-02-08 23:43:05,805 INFO [asr_datamodule.py:250] (1/4) Time warp factor: 80 +2023-02-08 23:43:05,805 INFO [asr_datamodule.py:260] (1/4) Num frame mask: 10 +2023-02-08 23:43:05,805 INFO [asr_datamodule.py:273] (1/4) About to create train dataset +2023-02-08 23:43:05,806 INFO [asr_datamodule.py:300] (1/4) Using DynamicBucketingSampler. +2023-02-08 23:43:05,825 WARNING [train.py:1067] (1/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-08 23:43:07,885 INFO [asr_datamodule.py:316] (1/4) About to create train dataloader +2023-02-08 23:43:07,885 INFO [asr_datamodule.py:430] (1/4) About to get dev-clean cuts +2023-02-08 23:43:07,910 INFO [asr_datamodule.py:437] (1/4) About to get dev-other cuts +2023-02-08 23:43:07,932 INFO [asr_datamodule.py:347] (1/4) About to create dev dataset +2023-02-08 23:43:08,281 INFO [asr_datamodule.py:364] (1/4) About to create dev dataloader +2023-02-08 23:43:08,281 INFO [train.py:1122] (1/4) Loading grad scaler state dict +2023-02-08 23:43:20,414 WARNING [train.py:1067] (1/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-08 23:43:25,772 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-02-08 23:43:26,066 INFO [train.py:901] (1/4) Epoch 28, batch 0, loss[loss=0.2683, simple_loss=0.3408, pruned_loss=0.09788, over 8194.00 frames. ], tot_loss[loss=0.2683, simple_loss=0.3408, pruned_loss=0.09788, over 8194.00 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:43:26,066 INFO [train.py:926] (1/4) Computing validation loss +2023-02-08 23:43:38,187 INFO [train.py:935] (1/4) Epoch 28, validation: loss=0.1714, simple_loss=0.2712, pruned_loss=0.03579, over 944034.00 frames. +2023-02-08 23:43:38,189 INFO [train.py:936] (1/4) Maximum memory allocated so far is 5898MB +2023-02-08 23:43:48,585 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218250.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:43:59,122 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-02-08 23:43:59,194 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218260.0, num_to_drop=1, layers_to_drop={0} +2023-02-08 23:44:26,822 INFO [train.py:901] (1/4) Epoch 28, batch 50, loss[loss=0.1753, simple_loss=0.2634, pruned_loss=0.04356, over 8243.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2837, pruned_loss=0.05927, over 366290.63 frames. ], batch size: 22, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:44:45,178 WARNING [train.py:1067] (1/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-02-08 23:44:46,084 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7173, 2.6415, 1.8021, 2.3582, 2.3274, 1.5716, 2.2139, 2.3342], + device='cuda:1'), covar=tensor([0.1510, 0.0464, 0.1334, 0.0670, 0.0686, 0.1713, 0.0987, 0.1043], + device='cuda:1'), in_proj_covar=tensor([0.0357, 0.0240, 0.0340, 0.0311, 0.0303, 0.0345, 0.0348, 0.0321], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-08 23:44:48,212 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.737e+02 2.503e+02 3.099e+02 3.838e+02 3.677e+03, threshold=6.198e+02, percent-clipped=7.0 +2023-02-08 23:45:01,314 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-02-08 23:45:09,775 INFO [train.py:901] (1/4) Epoch 28, batch 100, loss[loss=0.204, simple_loss=0.296, pruned_loss=0.05604, over 8363.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2829, pruned_loss=0.05853, over 642385.34 frames. ], batch size: 24, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:45:12,270 WARNING [train.py:1067] (1/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-02-08 23:45:42,208 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218375.0, num_to_drop=1, layers_to_drop={1} +2023-02-08 23:45:52,957 INFO [train.py:901] (1/4) Epoch 28, batch 150, loss[loss=0.2282, simple_loss=0.3142, pruned_loss=0.07105, over 8648.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2838, pruned_loss=0.05914, over 858455.84 frames. ], batch size: 48, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:46:01,151 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218397.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:46:01,340 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.75 vs. limit=5.0 +2023-02-08 23:46:10,824 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2140, 1.7323, 1.8397, 1.6457, 1.2663, 1.7026, 2.0329, 1.8486], + device='cuda:1'), covar=tensor([0.0560, 0.1178, 0.1600, 0.1380, 0.0582, 0.1404, 0.0639, 0.0606], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0161, 0.0101, 0.0163, 0.0112, 0.0146], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-08 23:46:12,821 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.615e+02 2.274e+02 2.796e+02 3.416e+02 5.816e+02, threshold=5.591e+02, percent-clipped=0.0 +2023-02-08 23:46:19,712 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218422.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:46:32,311 INFO [train.py:901] (1/4) Epoch 28, batch 200, loss[loss=0.1961, simple_loss=0.2857, pruned_loss=0.05328, over 8462.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.2831, pruned_loss=0.05882, over 1026204.14 frames. ], batch size: 25, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:46:50,626 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218462.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:47:10,724 INFO [train.py:901] (1/4) Epoch 28, batch 250, loss[loss=0.2152, simple_loss=0.2936, pruned_loss=0.06837, over 8369.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.283, pruned_loss=0.05827, over 1158780.76 frames. ], batch size: 24, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:47:23,080 WARNING [train.py:1067] (1/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-02-08 23:47:31,296 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.545e+02 2.405e+02 2.917e+02 3.543e+02 7.929e+02, threshold=5.833e+02, percent-clipped=6.0 +2023-02-08 23:47:33,450 WARNING [train.py:1067] (1/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-02-08 23:47:41,359 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=218527.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:47:48,876 INFO [train.py:901] (1/4) Epoch 28, batch 300, loss[loss=0.1857, simple_loss=0.2791, pruned_loss=0.04613, over 8472.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2834, pruned_loss=0.05837, over 1262917.92 frames. ], batch size: 25, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:47:53,401 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218544.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:14,318 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218572.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:18,152 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218577.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:25,696 INFO [train.py:901] (1/4) Epoch 28, batch 350, loss[loss=0.2123, simple_loss=0.2819, pruned_loss=0.07134, over 7539.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2833, pruned_loss=0.05854, over 1340446.29 frames. ], batch size: 18, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:48:28,669 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=218592.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:29,467 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7931, 1.8094, 2.7621, 2.1656, 2.4975, 1.8807, 1.6488, 1.4396], + device='cuda:1'), covar=tensor([0.7822, 0.6316, 0.2247, 0.4567, 0.3559, 0.4729, 0.3172, 0.5885], + device='cuda:1'), in_proj_covar=tensor([0.0959, 0.1017, 0.0823, 0.0986, 0.1018, 0.0922, 0.0763, 0.0846], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-08 23:48:43,884 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.664e+02 2.330e+02 2.853e+02 3.797e+02 9.826e+02, threshold=5.707e+02, percent-clipped=4.0 +2023-02-08 23:48:59,283 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218631.0, num_to_drop=1, layers_to_drop={0} +2023-02-08 23:49:04,797 INFO [train.py:901] (1/4) Epoch 28, batch 400, loss[loss=0.1695, simple_loss=0.2504, pruned_loss=0.0443, over 7546.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2837, pruned_loss=0.05845, over 1401020.76 frames. ], batch size: 18, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:49:16,380 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5696, 1.8973, 2.8968, 1.4347, 2.0187, 1.9120, 1.6802, 2.1809], + device='cuda:1'), covar=tensor([0.2055, 0.2812, 0.0889, 0.4805, 0.2163, 0.3463, 0.2544, 0.2484], + device='cuda:1'), in_proj_covar=tensor([0.0542, 0.0637, 0.0566, 0.0672, 0.0660, 0.0615, 0.0568, 0.0648], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-08 23:49:17,842 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218656.0, num_to_drop=1, layers_to_drop={1} +2023-02-08 23:49:19,964 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218659.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:49:40,564 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218687.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:49:41,126 INFO [train.py:901] (1/4) Epoch 28, batch 450, loss[loss=0.1768, simple_loss=0.269, pruned_loss=0.04226, over 8138.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.285, pruned_loss=0.05896, over 1452349.09 frames. ], batch size: 22, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:49:59,791 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.362e+02 2.836e+02 3.643e+02 9.062e+02, threshold=5.672e+02, percent-clipped=2.0 +2023-02-08 23:50:18,541 INFO [train.py:901] (1/4) Epoch 28, batch 500, loss[loss=0.174, simple_loss=0.2656, pruned_loss=0.04113, over 8479.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.284, pruned_loss=0.05823, over 1490572.80 frames. ], batch size: 27, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:50:57,129 INFO [train.py:901] (1/4) Epoch 28, batch 550, loss[loss=0.1905, simple_loss=0.2721, pruned_loss=0.05442, over 7928.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2846, pruned_loss=0.05871, over 1517920.14 frames. ], batch size: 20, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:51:05,247 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4076, 2.3219, 3.0593, 2.5460, 3.0690, 2.5139, 2.3124, 1.7083], + device='cuda:1'), covar=tensor([0.6236, 0.5386, 0.2420, 0.4133, 0.2828, 0.3352, 0.2121, 0.6004], + device='cuda:1'), in_proj_covar=tensor([0.0963, 0.1021, 0.0829, 0.0991, 0.1024, 0.0928, 0.0768, 0.0849], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-08 23:51:16,044 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.820e+02 2.392e+02 2.925e+02 3.560e+02 1.211e+03, threshold=5.850e+02, percent-clipped=4.0 +2023-02-08 23:51:29,357 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1711, 3.4918, 2.3107, 2.9802, 2.8350, 2.1765, 2.6837, 2.9419], + device='cuda:1'), covar=tensor([0.1682, 0.0415, 0.1238, 0.0789, 0.0818, 0.1523, 0.1278, 0.1230], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0244, 0.0342, 0.0315, 0.0305, 0.0349, 0.0352, 0.0325], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-08 23:51:30,152 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218833.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:51:33,469 INFO [train.py:901] (1/4) Epoch 28, batch 600, loss[loss=0.184, simple_loss=0.2694, pruned_loss=0.0493, over 8236.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2845, pruned_loss=0.05846, over 1541127.97 frames. ], batch size: 22, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:51:33,892 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.17 vs. limit=5.0 +2023-02-08 23:51:53,184 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218858.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:51:56,576 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-02-08 23:52:04,137 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218871.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:52:06,122 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=4.05 vs. limit=5.0 +2023-02-08 23:52:10,229 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3698, 1.6333, 1.6217, 1.1148, 1.6769, 1.3299, 0.3001, 1.5929], + device='cuda:1'), covar=tensor([0.0621, 0.0430, 0.0356, 0.0593, 0.0446, 0.1029, 0.1016, 0.0308], + device='cuda:1'), in_proj_covar=tensor([0.0474, 0.0412, 0.0366, 0.0458, 0.0395, 0.0553, 0.0402, 0.0442], + device='cuda:1'), out_proj_covar=tensor([1.2539e-04, 1.0696e-04, 9.5266e-05, 1.1974e-04, 1.0344e-04, 1.5414e-04, + 1.0732e-04, 1.1585e-04], device='cuda:1') +2023-02-08 23:52:18,545 INFO [train.py:901] (1/4) Epoch 28, batch 650, loss[loss=0.1632, simple_loss=0.2592, pruned_loss=0.03357, over 8502.00 frames. ], tot_loss[loss=0.199, simple_loss=0.283, pruned_loss=0.05745, over 1557046.85 frames. ], batch size: 26, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:52:40,036 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.733e+02 2.221e+02 2.637e+02 3.403e+02 7.509e+02, threshold=5.274e+02, percent-clipped=1.0 +2023-02-08 23:52:41,041 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218915.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:52:54,627 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.6974, 4.7304, 4.2054, 2.2003, 4.1287, 4.3007, 4.2105, 4.1061], + device='cuda:1'), covar=tensor([0.0586, 0.0468, 0.0951, 0.4330, 0.0844, 0.0885, 0.1253, 0.0765], + device='cuda:1'), in_proj_covar=tensor([0.0540, 0.0461, 0.0445, 0.0555, 0.0444, 0.0463, 0.0439, 0.0406], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-08 23:52:55,955 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218936.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:52:57,371 INFO [train.py:901] (1/4) Epoch 28, batch 700, loss[loss=0.1813, simple_loss=0.2578, pruned_loss=0.05242, over 7544.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.283, pruned_loss=0.05793, over 1567736.63 frames. ], batch size: 18, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:52:59,037 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218940.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:01,198 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218943.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:18,855 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218968.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:31,040 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=218983.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:33,178 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218986.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:34,534 INFO [train.py:901] (1/4) Epoch 28, batch 750, loss[loss=0.1927, simple_loss=0.2686, pruned_loss=0.05835, over 7545.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2837, pruned_loss=0.05842, over 1577188.68 frames. ], batch size: 18, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:53:46,816 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219002.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:55,126 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.438e+02 2.280e+02 2.810e+02 3.388e+02 7.203e+02, threshold=5.620e+02, percent-clipped=6.0 +2023-02-08 23:53:55,158 WARNING [train.py:1067] (1/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-02-08 23:54:04,597 WARNING [train.py:1067] (1/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-02-08 23:54:12,570 INFO [train.py:901] (1/4) Epoch 28, batch 800, loss[loss=0.1816, simple_loss=0.2669, pruned_loss=0.04815, over 7766.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2845, pruned_loss=0.05868, over 1580938.24 frames. ], batch size: 19, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:54:12,691 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.7407, 5.8447, 5.1602, 2.5476, 5.0919, 5.5450, 5.3198, 5.3865], + device='cuda:1'), covar=tensor([0.0503, 0.0426, 0.0913, 0.4093, 0.0803, 0.0730, 0.1134, 0.0585], + device='cuda:1'), in_proj_covar=tensor([0.0541, 0.0460, 0.0445, 0.0555, 0.0444, 0.0464, 0.0438, 0.0405], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-08 23:54:13,416 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219039.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:54:22,016 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219051.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:54:49,148 INFO [train.py:901] (1/4) Epoch 28, batch 850, loss[loss=0.1716, simple_loss=0.2641, pruned_loss=0.03951, over 8294.00 frames. ], tot_loss[loss=0.201, simple_loss=0.285, pruned_loss=0.05857, over 1594305.23 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:54:50,761 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219090.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:55:10,252 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.529e+02 2.432e+02 3.183e+02 3.929e+02 8.024e+02, threshold=6.365e+02, percent-clipped=6.0 +2023-02-08 23:55:11,241 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4880, 2.3518, 3.1349, 2.5101, 2.9512, 2.6153, 2.4291, 1.9589], + device='cuda:1'), covar=tensor([0.5672, 0.5160, 0.1918, 0.3882, 0.2629, 0.3065, 0.1945, 0.5452], + device='cuda:1'), in_proj_covar=tensor([0.0965, 0.1023, 0.0830, 0.0991, 0.1022, 0.0928, 0.0769, 0.0848], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-08 23:55:15,816 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-08 23:55:27,539 INFO [train.py:901] (1/4) Epoch 28, batch 900, loss[loss=0.2437, simple_loss=0.3295, pruned_loss=0.07898, over 8636.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2844, pruned_loss=0.05882, over 1601735.97 frames. ], batch size: 31, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:56:03,870 INFO [train.py:901] (1/4) Epoch 28, batch 950, loss[loss=0.1828, simple_loss=0.2698, pruned_loss=0.04792, over 8291.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2838, pruned_loss=0.05843, over 1605752.73 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:56:15,303 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.3571, 1.6068, 4.5913, 1.5764, 4.0353, 3.8244, 4.1084, 4.0109], + device='cuda:1'), covar=tensor([0.0639, 0.4558, 0.0545, 0.4689, 0.1139, 0.1019, 0.0644, 0.0673], + device='cuda:1'), in_proj_covar=tensor([0.0676, 0.0659, 0.0729, 0.0654, 0.0739, 0.0631, 0.0637, 0.0711], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-08 23:56:22,890 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.715e+02 2.524e+02 3.053e+02 4.249e+02 9.516e+02, threshold=6.106e+02, percent-clipped=7.0 +2023-02-08 23:56:29,702 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219221.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:56:34,848 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-02-08 23:56:40,761 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.1144, 1.5782, 4.3410, 1.5518, 3.8073, 3.6311, 3.9262, 3.7924], + device='cuda:1'), covar=tensor([0.0821, 0.4518, 0.0605, 0.4531, 0.1263, 0.1001, 0.0678, 0.0770], + device='cuda:1'), in_proj_covar=tensor([0.0676, 0.0659, 0.0727, 0.0653, 0.0739, 0.0631, 0.0636, 0.0711], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-08 23:56:43,596 INFO [train.py:901] (1/4) Epoch 28, batch 1000, loss[loss=0.1735, simple_loss=0.2642, pruned_loss=0.0414, over 8196.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2833, pruned_loss=0.0586, over 1605568.02 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:56:46,608 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219242.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:04,452 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219267.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:11,554 WARNING [train.py:1067] (1/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-02-08 23:57:19,431 INFO [train.py:901] (1/4) Epoch 28, batch 1050, loss[loss=0.2094, simple_loss=0.3056, pruned_loss=0.05661, over 8345.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.283, pruned_loss=0.05799, over 1609162.29 frames. ], batch size: 26, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:57:23,672 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-02-08 23:57:33,307 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219307.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:38,278 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.456e+02 2.957e+02 3.788e+02 8.190e+02, threshold=5.915e+02, percent-clipped=1.0 +2023-02-08 23:57:47,811 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219327.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:52,097 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219332.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:52,298 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-02-08 23:57:56,808 INFO [train.py:901] (1/4) Epoch 28, batch 1100, loss[loss=0.1987, simple_loss=0.2831, pruned_loss=0.05716, over 7407.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2841, pruned_loss=0.05842, over 1614870.02 frames. ], batch size: 17, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:58:03,512 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219346.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:58:13,591 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219360.0, num_to_drop=1, layers_to_drop={0} +2023-02-08 23:58:30,134 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219383.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:58:33,473 INFO [train.py:901] (1/4) Epoch 28, batch 1150, loss[loss=0.2219, simple_loss=0.2858, pruned_loss=0.079, over 5956.00 frames. ], tot_loss[loss=0.199, simple_loss=0.283, pruned_loss=0.05747, over 1616299.61 frames. ], batch size: 13, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:58:37,163 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-02-08 23:58:52,532 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.530e+02 2.386e+02 3.071e+02 3.782e+02 1.293e+03, threshold=6.141e+02, percent-clipped=2.0 +2023-02-08 23:59:07,138 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219434.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:59:10,012 INFO [train.py:901] (1/4) Epoch 28, batch 1200, loss[loss=0.1741, simple_loss=0.252, pruned_loss=0.04812, over 7817.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2824, pruned_loss=0.0571, over 1614923.54 frames. ], batch size: 20, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:59:13,031 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219442.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:59:27,991 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219461.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:59:47,588 INFO [train.py:901] (1/4) Epoch 28, batch 1250, loss[loss=0.1883, simple_loss=0.2865, pruned_loss=0.04502, over 8313.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2831, pruned_loss=0.05744, over 1617371.50 frames. ], batch size: 25, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:59:48,512 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.7991, 1.7608, 1.9405, 1.7749, 0.9464, 1.6197, 2.2662, 2.3378], + device='cuda:1'), covar=tensor([0.0457, 0.1228, 0.1640, 0.1395, 0.0609, 0.1522, 0.0612, 0.0521], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0188, 0.0161, 0.0101, 0.0163, 0.0112, 0.0146], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-08 23:59:55,027 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219498.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:00:02,350 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219508.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:00:06,612 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.598e+02 2.357e+02 2.809e+02 3.466e+02 7.121e+02, threshold=5.618e+02, percent-clipped=3.0 +2023-02-09 00:00:23,407 INFO [train.py:901] (1/4) Epoch 28, batch 1300, loss[loss=0.1791, simple_loss=0.2596, pruned_loss=0.04932, over 7450.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2826, pruned_loss=0.05742, over 1617401.36 frames. ], batch size: 17, lr: 2.71e-03, grad_scale: 8.0 +2023-02-09 00:00:25,004 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9567, 2.0423, 1.7615, 2.7417, 1.3098, 1.5590, 1.9620, 2.0295], + device='cuda:1'), covar=tensor([0.0724, 0.0826, 0.0895, 0.0385, 0.1039, 0.1318, 0.0821, 0.0854], + device='cuda:1'), in_proj_covar=tensor([0.0227, 0.0191, 0.0240, 0.0209, 0.0199, 0.0242, 0.0245, 0.0202], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 00:00:31,519 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219549.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:00:42,920 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219565.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:01:02,333 INFO [train.py:901] (1/4) Epoch 28, batch 1350, loss[loss=0.1871, simple_loss=0.2701, pruned_loss=0.05202, over 7805.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2835, pruned_loss=0.0579, over 1619426.85 frames. ], batch size: 20, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:01:22,005 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.513e+02 2.365e+02 2.856e+02 3.377e+02 7.819e+02, threshold=5.713e+02, percent-clipped=4.0 +2023-02-09 00:01:39,663 INFO [train.py:901] (1/4) Epoch 28, batch 1400, loss[loss=0.1699, simple_loss=0.2609, pruned_loss=0.03942, over 8032.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2824, pruned_loss=0.05724, over 1618286.34 frames. ], batch size: 22, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:02:09,798 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219680.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:15,336 INFO [train.py:901] (1/4) Epoch 28, batch 1450, loss[loss=0.1869, simple_loss=0.2796, pruned_loss=0.04708, over 7962.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2814, pruned_loss=0.05694, over 1617854.88 frames. ], batch size: 21, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:02:23,473 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219698.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:25,414 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-02-09 00:02:28,352 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219704.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:02:33,319 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.7187, 5.8762, 5.0332, 2.8170, 5.1107, 5.5225, 5.2685, 5.3230], + device='cuda:1'), covar=tensor([0.0513, 0.0302, 0.0806, 0.3771, 0.0797, 0.0905, 0.1025, 0.0647], + device='cuda:1'), in_proj_covar=tensor([0.0540, 0.0460, 0.0448, 0.0555, 0.0444, 0.0465, 0.0439, 0.0406], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:02:36,724 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.725e+02 2.298e+02 2.874e+02 3.536e+02 7.746e+02, threshold=5.748e+02, percent-clipped=3.0 +2023-02-09 00:02:39,218 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219717.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:43,524 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219723.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:54,214 INFO [train.py:901] (1/4) Epoch 28, batch 1500, loss[loss=0.1445, simple_loss=0.2279, pruned_loss=0.03056, over 7661.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2816, pruned_loss=0.05679, over 1616305.37 frames. ], batch size: 19, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:02:57,276 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219742.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:05,681 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219754.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:23,841 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219779.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:25,269 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219781.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:29,879 INFO [train.py:901] (1/4) Epoch 28, batch 1550, loss[loss=0.1752, simple_loss=0.2495, pruned_loss=0.05039, over 7289.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2823, pruned_loss=0.05687, over 1611910.38 frames. ], batch size: 16, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:03:42,605 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219805.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:49,435 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.606e+02 2.435e+02 2.945e+02 3.827e+02 6.900e+02, threshold=5.889e+02, percent-clipped=4.0 +2023-02-09 00:03:54,630 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219819.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:03:55,263 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6912, 1.4363, 1.6250, 1.3552, 0.9637, 1.4316, 1.5694, 1.2620], + device='cuda:1'), covar=tensor([0.0620, 0.1349, 0.1737, 0.1582, 0.0657, 0.1549, 0.0754, 0.0768], + device='cuda:1'), in_proj_covar=tensor([0.0098, 0.0152, 0.0188, 0.0161, 0.0101, 0.0162, 0.0112, 0.0145], + device='cuda:1'), out_proj_covar=tensor([0.0006, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 00:04:03,136 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219830.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:04:08,596 INFO [train.py:901] (1/4) Epoch 28, batch 1600, loss[loss=0.1983, simple_loss=0.2852, pruned_loss=0.0557, over 8465.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2831, pruned_loss=0.05693, over 1618696.92 frames. ], batch size: 25, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:04:18,703 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219852.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:04:35,335 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219875.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:04:44,469 INFO [train.py:901] (1/4) Epoch 28, batch 1650, loss[loss=0.2518, simple_loss=0.3326, pruned_loss=0.08546, over 8660.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2831, pruned_loss=0.05671, over 1617238.94 frames. ], batch size: 39, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:05:02,686 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.357e+02 2.482e+02 2.898e+02 3.443e+02 5.647e+02, threshold=5.797e+02, percent-clipped=0.0 +2023-02-09 00:05:20,267 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219936.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:05:21,462 INFO [train.py:901] (1/4) Epoch 28, batch 1700, loss[loss=0.2103, simple_loss=0.2864, pruned_loss=0.06712, over 8599.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2829, pruned_loss=0.05695, over 1616611.69 frames. ], batch size: 31, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:05:39,120 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219961.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:05:43,285 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219967.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:05:57,778 INFO [train.py:901] (1/4) Epoch 28, batch 1750, loss[loss=0.1651, simple_loss=0.2531, pruned_loss=0.0385, over 7558.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2826, pruned_loss=0.05628, over 1617141.64 frames. ], batch size: 18, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:06:17,587 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.415e+02 2.339e+02 2.848e+02 3.606e+02 1.047e+03, threshold=5.695e+02, percent-clipped=4.0 +2023-02-09 00:06:34,450 INFO [train.py:901] (1/4) Epoch 28, batch 1800, loss[loss=0.1605, simple_loss=0.2625, pruned_loss=0.0292, over 8257.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2823, pruned_loss=0.05667, over 1615792.92 frames. ], batch size: 24, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:07:02,975 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220075.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 00:07:11,957 INFO [train.py:901] (1/4) Epoch 28, batch 1850, loss[loss=0.1954, simple_loss=0.2831, pruned_loss=0.05385, over 8340.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2832, pruned_loss=0.05738, over 1619743.24 frames. ], batch size: 26, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:07:20,513 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220100.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 00:07:23,908 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220105.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:07:30,304 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.818e+02 2.328e+02 2.682e+02 3.608e+02 8.535e+02, threshold=5.364e+02, percent-clipped=7.0 +2023-02-09 00:07:31,749 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220116.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:07:38,071 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220125.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:07:47,173 INFO [train.py:901] (1/4) Epoch 28, batch 1900, loss[loss=0.2061, simple_loss=0.2969, pruned_loss=0.05764, over 8455.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2823, pruned_loss=0.05727, over 1615359.26 frames. ], batch size: 25, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:08:19,360 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-02-09 00:08:25,760 INFO [train.py:901] (1/4) Epoch 28, batch 1950, loss[loss=0.261, simple_loss=0.3313, pruned_loss=0.09537, over 7113.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2822, pruned_loss=0.05748, over 1606342.87 frames. ], batch size: 71, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:08:32,984 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-02-09 00:08:44,716 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.518e+02 2.461e+02 2.916e+02 3.869e+02 7.609e+02, threshold=5.833e+02, percent-clipped=8.0 +2023-02-09 00:08:48,263 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220219.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:08:51,065 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220223.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:08:53,656 WARNING [train.py:1067] (1/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-02-09 00:09:01,373 INFO [train.py:901] (1/4) Epoch 28, batch 2000, loss[loss=0.1839, simple_loss=0.2746, pruned_loss=0.04655, over 8511.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2807, pruned_loss=0.05641, over 1604694.27 frames. ], batch size: 28, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:09:02,862 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220240.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:07,808 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220247.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:08,581 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220248.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:29,128 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220276.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:37,451 INFO [train.py:901] (1/4) Epoch 28, batch 2050, loss[loss=0.2185, simple_loss=0.3066, pruned_loss=0.0652, over 8505.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2815, pruned_loss=0.05618, over 1610568.54 frames. ], batch size: 28, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:09:58,200 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.689e+02 2.398e+02 2.757e+02 3.324e+02 6.340e+02, threshold=5.514e+02, percent-clipped=2.0 +2023-02-09 00:10:03,993 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9910, 1.8462, 2.2612, 1.9992, 2.2198, 2.1288, 1.9405, 1.1763], + device='cuda:1'), covar=tensor([0.6027, 0.4949, 0.2184, 0.4005, 0.2760, 0.3324, 0.1994, 0.5320], + device='cuda:1'), in_proj_covar=tensor([0.0970, 0.1029, 0.0835, 0.0997, 0.1028, 0.0934, 0.0773, 0.0853], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 00:10:12,766 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220334.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:10:15,436 INFO [train.py:901] (1/4) Epoch 28, batch 2100, loss[loss=0.1995, simple_loss=0.2803, pruned_loss=0.05933, over 8637.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.281, pruned_loss=0.05607, over 1614825.54 frames. ], batch size: 49, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:10:20,359 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-02-09 00:10:42,392 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.60 vs. limit=2.0 +2023-02-09 00:10:51,259 INFO [train.py:901] (1/4) Epoch 28, batch 2150, loss[loss=0.1655, simple_loss=0.2522, pruned_loss=0.03939, over 7449.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2821, pruned_loss=0.05708, over 1618847.65 frames. ], batch size: 17, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:11:11,484 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.499e+02 2.504e+02 2.973e+02 4.041e+02 1.001e+03, threshold=5.945e+02, percent-clipped=8.0 +2023-02-09 00:11:28,328 INFO [train.py:901] (1/4) Epoch 28, batch 2200, loss[loss=0.1991, simple_loss=0.2837, pruned_loss=0.05731, over 8338.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2812, pruned_loss=0.05699, over 1620495.91 frames. ], batch size: 26, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:11:36,288 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220449.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:11:44,017 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220460.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:12:03,395 INFO [train.py:901] (1/4) Epoch 28, batch 2250, loss[loss=0.2148, simple_loss=0.3033, pruned_loss=0.06315, over 8321.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2819, pruned_loss=0.05723, over 1621023.59 frames. ], batch size: 25, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:12:09,254 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220496.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:12:15,026 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.42 vs. limit=5.0 +2023-02-09 00:12:22,281 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.608e+02 2.331e+02 2.835e+02 3.325e+02 7.200e+02, threshold=5.671e+02, percent-clipped=3.0 +2023-02-09 00:12:27,560 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220521.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:12:41,520 INFO [train.py:901] (1/4) Epoch 28, batch 2300, loss[loss=0.1599, simple_loss=0.2435, pruned_loss=0.03819, over 7973.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2805, pruned_loss=0.05621, over 1622945.87 frames. ], batch size: 21, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:12:44,383 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.6690, 1.5456, 4.8827, 1.8673, 4.3126, 3.9983, 4.4269, 4.2989], + device='cuda:1'), covar=tensor([0.0597, 0.5088, 0.0485, 0.4448, 0.1143, 0.1031, 0.0584, 0.0675], + device='cuda:1'), in_proj_covar=tensor([0.0682, 0.0670, 0.0739, 0.0664, 0.0752, 0.0639, 0.0646, 0.0722], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:12:51,184 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2257, 1.1535, 2.1148, 1.1871, 2.0109, 2.2442, 2.3944, 1.9438], + device='cuda:1'), covar=tensor([0.1168, 0.1439, 0.0486, 0.1936, 0.0895, 0.0387, 0.0731, 0.0634], + device='cuda:1'), in_proj_covar=tensor([0.0308, 0.0327, 0.0293, 0.0321, 0.0323, 0.0277, 0.0441, 0.0308], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 00:12:59,738 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220564.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:06,020 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220573.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:07,454 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220575.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:16,608 INFO [train.py:901] (1/4) Epoch 28, batch 2350, loss[loss=0.1805, simple_loss=0.2717, pruned_loss=0.04465, over 8478.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2808, pruned_loss=0.05617, over 1624755.70 frames. ], batch size: 25, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:13:18,215 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220590.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:18,797 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220591.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:35,684 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.370e+02 2.329e+02 2.956e+02 3.826e+02 8.837e+02, threshold=5.912e+02, percent-clipped=4.0 +2023-02-09 00:13:36,669 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220615.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:40,249 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220620.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:53,609 INFO [train.py:901] (1/4) Epoch 28, batch 2400, loss[loss=0.2175, simple_loss=0.2837, pruned_loss=0.07566, over 7697.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2806, pruned_loss=0.05635, over 1622221.92 frames. ], batch size: 18, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:14:16,668 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220669.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:14:29,705 INFO [train.py:901] (1/4) Epoch 28, batch 2450, loss[loss=0.1975, simple_loss=0.2793, pruned_loss=0.05786, over 7989.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2801, pruned_loss=0.0562, over 1619836.04 frames. ], batch size: 21, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:14:42,736 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220706.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:14:48,793 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.600e+02 2.507e+02 3.309e+02 3.917e+02 8.053e+02, threshold=6.618e+02, percent-clipped=4.0 +2023-02-09 00:15:03,092 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220735.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:15:05,104 INFO [train.py:901] (1/4) Epoch 28, batch 2500, loss[loss=0.2054, simple_loss=0.2727, pruned_loss=0.06907, over 7797.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2801, pruned_loss=0.05612, over 1621520.13 frames. ], batch size: 20, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:15:15,098 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3933, 1.4261, 1.3815, 1.8062, 0.7363, 1.2280, 1.3276, 1.4626], + device='cuda:1'), covar=tensor([0.0836, 0.0774, 0.0950, 0.0464, 0.1112, 0.1357, 0.0689, 0.0731], + device='cuda:1'), in_proj_covar=tensor([0.0229, 0.0193, 0.0242, 0.0211, 0.0202, 0.0245, 0.0249, 0.0204], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 00:15:42,737 INFO [train.py:901] (1/4) Epoch 28, batch 2550, loss[loss=0.2284, simple_loss=0.3069, pruned_loss=0.07492, over 8033.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2809, pruned_loss=0.05691, over 1622148.72 frames. ], batch size: 22, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:16:02,764 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.451e+02 2.505e+02 3.011e+02 3.782e+02 1.017e+03, threshold=6.023e+02, percent-clipped=3.0 +2023-02-09 00:16:06,710 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220820.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:14,502 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220831.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:19,226 INFO [train.py:901] (1/4) Epoch 28, batch 2600, loss[loss=0.2093, simple_loss=0.2923, pruned_loss=0.06313, over 8638.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2817, pruned_loss=0.05753, over 1618730.13 frames. ], batch size: 34, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:16:20,649 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220840.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:24,294 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220845.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:32,235 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220856.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:57,368 INFO [train.py:901] (1/4) Epoch 28, batch 2650, loss[loss=0.1976, simple_loss=0.2817, pruned_loss=0.05672, over 8246.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2804, pruned_loss=0.05657, over 1619829.29 frames. ], batch size: 24, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:17:12,884 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5345, 2.4328, 1.7452, 2.1874, 2.1024, 1.5653, 2.0734, 2.1044], + device='cuda:1'), covar=tensor([0.1517, 0.0505, 0.1330, 0.0684, 0.0753, 0.1602, 0.1006, 0.1022], + device='cuda:1'), in_proj_covar=tensor([0.0356, 0.0242, 0.0339, 0.0311, 0.0301, 0.0345, 0.0346, 0.0321], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:17:16,289 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.526e+02 2.381e+02 2.801e+02 3.642e+02 5.464e+02, threshold=5.602e+02, percent-clipped=0.0 +2023-02-09 00:17:17,796 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220917.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:17:21,293 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220922.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:17:32,910 INFO [train.py:901] (1/4) Epoch 28, batch 2700, loss[loss=0.2108, simple_loss=0.3032, pruned_loss=0.05916, over 8264.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2811, pruned_loss=0.05685, over 1619055.77 frames. ], batch size: 24, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:17:37,481 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6231, 1.4215, 1.6771, 1.2868, 0.9104, 1.4221, 1.4750, 1.3853], + device='cuda:1'), covar=tensor([0.0619, 0.1250, 0.1615, 0.1559, 0.0610, 0.1461, 0.0742, 0.0687], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0162, 0.0102, 0.0163, 0.0113, 0.0146], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 00:17:50,477 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220962.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:09,043 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220987.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:09,535 INFO [train.py:901] (1/4) Epoch 28, batch 2750, loss[loss=0.1794, simple_loss=0.2699, pruned_loss=0.04445, over 8497.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2809, pruned_loss=0.05725, over 1614361.09 frames. ], batch size: 26, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:18:11,782 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220991.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:29,675 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221013.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:31,038 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.597e+02 2.419e+02 2.908e+02 3.517e+02 7.342e+02, threshold=5.816e+02, percent-clipped=5.0 +2023-02-09 00:18:31,965 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221016.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:43,261 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221032.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:47,244 INFO [train.py:901] (1/4) Epoch 28, batch 2800, loss[loss=0.1901, simple_loss=0.2792, pruned_loss=0.05052, over 8023.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2793, pruned_loss=0.05627, over 1607516.44 frames. ], batch size: 22, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:18:57,816 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8694, 1.7236, 2.3799, 1.5668, 1.5091, 2.2745, 0.4322, 1.4893], + device='cuda:1'), covar=tensor([0.1346, 0.1386, 0.0320, 0.0908, 0.2094, 0.0404, 0.1774, 0.1172], + device='cuda:1'), in_proj_covar=tensor([0.0201, 0.0205, 0.0136, 0.0224, 0.0277, 0.0146, 0.0174, 0.0199], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 00:19:22,673 INFO [train.py:901] (1/4) Epoch 28, batch 2850, loss[loss=0.1997, simple_loss=0.2813, pruned_loss=0.05902, over 7988.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2803, pruned_loss=0.05671, over 1607164.15 frames. ], batch size: 21, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:19:43,232 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.362e+02 2.366e+02 2.856e+02 3.627e+02 6.501e+02, threshold=5.713e+02, percent-clipped=2.0 +2023-02-09 00:19:53,962 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221128.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:20:00,677 INFO [train.py:901] (1/4) Epoch 28, batch 2900, loss[loss=0.209, simple_loss=0.3066, pruned_loss=0.05565, over 8103.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2815, pruned_loss=0.05747, over 1611841.38 frames. ], batch size: 23, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:20:04,292 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4066, 1.5967, 2.1386, 1.3625, 1.5570, 1.6746, 1.5080, 1.5194], + device='cuda:1'), covar=tensor([0.2119, 0.2826, 0.1015, 0.4961, 0.2137, 0.3683, 0.2703, 0.2295], + device='cuda:1'), in_proj_covar=tensor([0.0541, 0.0639, 0.0566, 0.0673, 0.0663, 0.0612, 0.0566, 0.0647], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:20:32,274 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-02-09 00:20:33,725 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221184.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:20:36,502 INFO [train.py:901] (1/4) Epoch 28, batch 2950, loss[loss=0.1738, simple_loss=0.2506, pruned_loss=0.04851, over 7783.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2826, pruned_loss=0.05815, over 1613526.16 frames. ], batch size: 19, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:20:43,328 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.0329, 1.1978, 1.5635, 1.0058, 1.1348, 1.2310, 1.1161, 1.1427], + device='cuda:1'), covar=tensor([0.1473, 0.2008, 0.0780, 0.3504, 0.1605, 0.2606, 0.1905, 0.2017], + device='cuda:1'), in_proj_covar=tensor([0.0541, 0.0640, 0.0567, 0.0674, 0.0663, 0.0612, 0.0567, 0.0648], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:20:55,456 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.514e+02 2.299e+02 2.993e+02 3.879e+02 1.208e+03, threshold=5.985e+02, percent-clipped=10.0 +2023-02-09 00:21:01,649 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=4.04 vs. limit=5.0 +2023-02-09 00:21:13,540 INFO [train.py:901] (1/4) Epoch 28, batch 3000, loss[loss=0.1673, simple_loss=0.2462, pruned_loss=0.04419, over 7923.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2812, pruned_loss=0.0575, over 1607359.71 frames. ], batch size: 20, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:21:13,541 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 00:21:31,972 INFO [train.py:935] (1/4) Epoch 28, validation: loss=0.1712, simple_loss=0.2708, pruned_loss=0.03578, over 944034.00 frames. +2023-02-09 00:21:31,973 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 00:21:47,605 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2137, 1.0783, 1.3128, 1.0302, 0.9604, 1.3275, 0.0885, 0.9492], + device='cuda:1'), covar=tensor([0.1516, 0.1226, 0.0499, 0.0708, 0.2366, 0.0554, 0.1891, 0.1131], + device='cuda:1'), in_proj_covar=tensor([0.0201, 0.0205, 0.0137, 0.0225, 0.0278, 0.0147, 0.0174, 0.0199], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 00:21:54,455 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221266.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:21:55,519 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-02-09 00:22:03,528 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.65 vs. limit=5.0 +2023-02-09 00:22:10,139 INFO [train.py:901] (1/4) Epoch 28, batch 3050, loss[loss=0.1868, simple_loss=0.2537, pruned_loss=0.05998, over 7703.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2828, pruned_loss=0.0582, over 1612634.16 frames. ], batch size: 18, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:22:10,359 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221288.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:22:18,082 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221299.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:22:28,217 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221313.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:22:29,357 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.568e+02 2.361e+02 2.830e+02 3.600e+02 1.199e+03, threshold=5.660e+02, percent-clipped=4.0 +2023-02-09 00:22:37,155 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.7967, 2.2239, 3.7181, 1.6816, 1.9645, 3.7088, 0.7593, 2.1232], + device='cuda:1'), covar=tensor([0.1424, 0.1097, 0.0195, 0.1649, 0.1986, 0.0220, 0.1851, 0.1184], + device='cuda:1'), in_proj_covar=tensor([0.0201, 0.0205, 0.0137, 0.0224, 0.0277, 0.0146, 0.0174, 0.0199], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 00:22:45,360 INFO [train.py:901] (1/4) Epoch 28, batch 3100, loss[loss=0.1815, simple_loss=0.2644, pruned_loss=0.04928, over 7527.00 frames. ], tot_loss[loss=0.2, simple_loss=0.283, pruned_loss=0.0585, over 1613693.17 frames. ], batch size: 18, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:23:18,456 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221381.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:23:20,619 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221384.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:23:23,254 INFO [train.py:901] (1/4) Epoch 28, batch 3150, loss[loss=0.1941, simple_loss=0.2698, pruned_loss=0.0592, over 7217.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2821, pruned_loss=0.0578, over 1612835.35 frames. ], batch size: 16, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:23:31,492 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9000, 1.5283, 3.1493, 1.5252, 2.3499, 3.3569, 3.5257, 2.8851], + device='cuda:1'), covar=tensor([0.1215, 0.1797, 0.0328, 0.2115, 0.0979, 0.0270, 0.0520, 0.0546], + device='cuda:1'), in_proj_covar=tensor([0.0308, 0.0329, 0.0294, 0.0323, 0.0326, 0.0278, 0.0443, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 00:23:34,610 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5038, 2.3821, 1.8393, 2.2864, 2.0388, 1.5431, 2.0329, 2.2461], + device='cuda:1'), covar=tensor([0.1621, 0.0660, 0.1592, 0.0668, 0.0908, 0.2036, 0.1199, 0.0999], + device='cuda:1'), in_proj_covar=tensor([0.0362, 0.0246, 0.0344, 0.0315, 0.0305, 0.0350, 0.0351, 0.0325], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:23:38,987 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221409.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:23:43,026 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.782e+02 2.344e+02 3.031e+02 3.872e+02 9.124e+02, threshold=6.062e+02, percent-clipped=5.0 +2023-02-09 00:24:00,287 INFO [train.py:901] (1/4) Epoch 28, batch 3200, loss[loss=0.1826, simple_loss=0.2722, pruned_loss=0.04647, over 8440.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2821, pruned_loss=0.05778, over 1610285.09 frames. ], batch size: 29, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:24:21,674 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0606, 1.5746, 3.5073, 1.7013, 2.6178, 3.8521, 3.9248, 3.3329], + device='cuda:1'), covar=tensor([0.1115, 0.1824, 0.0307, 0.1928, 0.0939, 0.0219, 0.0613, 0.0543], + device='cuda:1'), in_proj_covar=tensor([0.0308, 0.0329, 0.0294, 0.0323, 0.0325, 0.0278, 0.0442, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 00:24:36,745 INFO [train.py:901] (1/4) Epoch 28, batch 3250, loss[loss=0.1887, simple_loss=0.2546, pruned_loss=0.06138, over 7218.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.281, pruned_loss=0.05709, over 1606491.85 frames. ], batch size: 16, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:24:56,686 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.359e+02 2.800e+02 3.771e+02 8.910e+02, threshold=5.600e+02, percent-clipped=3.0 +2023-02-09 00:25:04,045 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=221525.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:25:12,972 INFO [train.py:901] (1/4) Epoch 28, batch 3300, loss[loss=0.1888, simple_loss=0.2817, pruned_loss=0.04798, over 8697.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2807, pruned_loss=0.05683, over 1605824.85 frames. ], batch size: 39, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:25:23,656 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6337, 1.9780, 3.0082, 1.4662, 2.3178, 2.0464, 1.7151, 2.3745], + device='cuda:1'), covar=tensor([0.1940, 0.2849, 0.0860, 0.4889, 0.1838, 0.3286, 0.2559, 0.2117], + device='cuda:1'), in_proj_covar=tensor([0.0536, 0.0634, 0.0561, 0.0668, 0.0657, 0.0605, 0.0561, 0.0641], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:25:25,041 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221555.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:25:42,943 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221580.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:25:48,454 INFO [train.py:901] (1/4) Epoch 28, batch 3350, loss[loss=0.1783, simple_loss=0.2653, pruned_loss=0.04566, over 8341.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2814, pruned_loss=0.05653, over 1614886.61 frames. ], batch size: 26, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:25:51,504 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3730, 2.5094, 2.0939, 4.0300, 1.5370, 1.8806, 2.4098, 2.7962], + device='cuda:1'), covar=tensor([0.0763, 0.0889, 0.0952, 0.0236, 0.1183, 0.1332, 0.0999, 0.0797], + device='cuda:1'), in_proj_covar=tensor([0.0230, 0.0194, 0.0243, 0.0212, 0.0202, 0.0245, 0.0250, 0.0204], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 00:26:09,958 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.399e+02 2.531e+02 3.062e+02 3.663e+02 8.444e+02, threshold=6.124e+02, percent-clipped=3.0 +2023-02-09 00:26:26,095 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221637.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:26:26,629 INFO [train.py:901] (1/4) Epoch 28, batch 3400, loss[loss=0.1955, simple_loss=0.2915, pruned_loss=0.0498, over 8254.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2822, pruned_loss=0.05699, over 1620452.72 frames. ], batch size: 24, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:26:29,006 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5744, 1.3229, 2.4160, 1.3511, 2.2285, 2.5786, 2.7376, 2.2162], + device='cuda:1'), covar=tensor([0.1057, 0.1491, 0.0382, 0.2027, 0.0752, 0.0369, 0.0664, 0.0616], + device='cuda:1'), in_proj_covar=tensor([0.0308, 0.0329, 0.0294, 0.0323, 0.0325, 0.0278, 0.0442, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 00:26:36,014 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0394, 1.6035, 1.7695, 1.4126, 0.9441, 1.5015, 1.8135, 1.5182], + device='cuda:1'), covar=tensor([0.0549, 0.1235, 0.1626, 0.1473, 0.0607, 0.1552, 0.0711, 0.0689], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0190, 0.0163, 0.0102, 0.0163, 0.0113, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 00:26:43,876 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221662.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:27:02,341 INFO [train.py:901] (1/4) Epoch 28, batch 3450, loss[loss=0.236, simple_loss=0.3239, pruned_loss=0.07402, over 8501.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.282, pruned_loss=0.05715, over 1619167.76 frames. ], batch size: 26, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:27:20,948 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7083, 2.3446, 3.9722, 1.6216, 3.0908, 2.3490, 1.8767, 2.9512], + device='cuda:1'), covar=tensor([0.1929, 0.2715, 0.0875, 0.4557, 0.1643, 0.3250, 0.2409, 0.2327], + device='cuda:1'), in_proj_covar=tensor([0.0537, 0.0635, 0.0563, 0.0670, 0.0660, 0.0608, 0.0562, 0.0644], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:27:21,422 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.306e+02 2.763e+02 3.583e+02 8.756e+02, threshold=5.526e+02, percent-clipped=3.0 +2023-02-09 00:27:39,500 INFO [train.py:901] (1/4) Epoch 28, batch 3500, loss[loss=0.2059, simple_loss=0.2778, pruned_loss=0.06702, over 7518.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2831, pruned_loss=0.05781, over 1620207.11 frames. ], batch size: 18, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:27:42,565 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-02-09 00:28:03,519 WARNING [train.py:1067] (1/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-02-09 00:28:07,250 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=221776.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:28:15,816 INFO [train.py:901] (1/4) Epoch 28, batch 3550, loss[loss=0.2092, simple_loss=0.2874, pruned_loss=0.06549, over 8467.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2822, pruned_loss=0.05726, over 1621237.60 frames. ], batch size: 25, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:28:35,272 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.790e+02 2.405e+02 2.949e+02 3.672e+02 8.337e+02, threshold=5.897e+02, percent-clipped=3.0 +2023-02-09 00:28:36,052 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.8400, 3.7830, 3.4334, 1.8744, 3.4122, 3.5386, 3.3277, 3.3814], + device='cuda:1'), covar=tensor([0.0880, 0.0671, 0.1262, 0.5060, 0.1046, 0.1071, 0.1633, 0.0861], + device='cuda:1'), in_proj_covar=tensor([0.0542, 0.0457, 0.0448, 0.0560, 0.0444, 0.0465, 0.0441, 0.0406], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:28:52,587 INFO [train.py:901] (1/4) Epoch 28, batch 3600, loss[loss=0.2194, simple_loss=0.2988, pruned_loss=0.07, over 8280.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2827, pruned_loss=0.05788, over 1617634.92 frames. ], batch size: 23, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:29:07,255 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6132, 1.3976, 1.6110, 1.2939, 0.9420, 1.4198, 1.5008, 1.3981], + device='cuda:1'), covar=tensor([0.0603, 0.1278, 0.1712, 0.1545, 0.0612, 0.1495, 0.0718, 0.0674], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0161, 0.0101, 0.0162, 0.0112, 0.0146], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 00:29:08,665 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7985, 2.2619, 3.8612, 1.6707, 2.8742, 2.3096, 1.9083, 2.9688], + device='cuda:1'), covar=tensor([0.1949, 0.2826, 0.0790, 0.4753, 0.1923, 0.3256, 0.2505, 0.2283], + device='cuda:1'), in_proj_covar=tensor([0.0536, 0.0634, 0.0563, 0.0670, 0.0659, 0.0607, 0.0561, 0.0642], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:29:15,450 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221869.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:29:28,481 INFO [train.py:901] (1/4) Epoch 28, batch 3650, loss[loss=0.1602, simple_loss=0.2516, pruned_loss=0.03441, over 7948.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2826, pruned_loss=0.05794, over 1613840.82 frames. ], batch size: 20, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:29:42,354 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=221908.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:29:47,072 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.635e+02 2.399e+02 3.022e+02 3.885e+02 8.966e+02, threshold=6.044e+02, percent-clipped=2.0 +2023-02-09 00:29:53,450 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0245, 2.5395, 2.6110, 1.5260, 2.8696, 1.8116, 1.5200, 2.1281], + device='cuda:1'), covar=tensor([0.1217, 0.0502, 0.0512, 0.1193, 0.0682, 0.1181, 0.1282, 0.0720], + device='cuda:1'), in_proj_covar=tensor([0.0470, 0.0409, 0.0362, 0.0458, 0.0393, 0.0548, 0.0401, 0.0440], + device='cuda:1'), out_proj_covar=tensor([1.2435e-04, 1.0588e-04, 9.4179e-05, 1.1959e-04, 1.0268e-04, 1.5285e-04, + 1.0707e-04, 1.1512e-04], device='cuda:1') +2023-02-09 00:29:54,449 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-02-09 00:29:57,465 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.9869, 1.7595, 6.1676, 2.4403, 5.5986, 5.1184, 5.6715, 5.6340], + device='cuda:1'), covar=tensor([0.0469, 0.4849, 0.0361, 0.3793, 0.0890, 0.0895, 0.0475, 0.0465], + device='cuda:1'), in_proj_covar=tensor([0.0681, 0.0670, 0.0737, 0.0662, 0.0748, 0.0637, 0.0645, 0.0718], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:30:02,963 INFO [train.py:901] (1/4) Epoch 28, batch 3700, loss[loss=0.1969, simple_loss=0.2846, pruned_loss=0.05457, over 8641.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2826, pruned_loss=0.05812, over 1611447.56 frames. ], batch size: 27, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:30:05,070 WARNING [train.py:1067] (1/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-02-09 00:30:38,683 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221984.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:30:41,396 INFO [train.py:901] (1/4) Epoch 28, batch 3750, loss[loss=0.1794, simple_loss=0.277, pruned_loss=0.04085, over 8105.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2833, pruned_loss=0.05812, over 1608756.08 frames. ], batch size: 23, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:31:01,419 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.685e+02 2.662e+02 3.142e+02 4.083e+02 1.270e+03, threshold=6.284e+02, percent-clipped=8.0 +2023-02-09 00:31:04,444 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3661, 1.9851, 4.0061, 1.9487, 2.6106, 4.5899, 4.6994, 3.9494], + device='cuda:1'), covar=tensor([0.1136, 0.1756, 0.0347, 0.2022, 0.1265, 0.0196, 0.0363, 0.0542], + device='cuda:1'), in_proj_covar=tensor([0.0310, 0.0330, 0.0295, 0.0325, 0.0326, 0.0279, 0.0445, 0.0311], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 00:31:10,190 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222027.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:31:17,962 INFO [train.py:901] (1/4) Epoch 28, batch 3800, loss[loss=0.2121, simple_loss=0.2766, pruned_loss=0.0738, over 7796.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2833, pruned_loss=0.05813, over 1613365.27 frames. ], batch size: 19, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:31:55,279 INFO [train.py:901] (1/4) Epoch 28, batch 3850, loss[loss=0.1995, simple_loss=0.2922, pruned_loss=0.05338, over 8365.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2838, pruned_loss=0.0582, over 1613745.73 frames. ], batch size: 24, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:32:11,759 WARNING [train.py:1067] (1/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-02-09 00:32:13,842 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.497e+02 2.208e+02 2.768e+02 3.453e+02 7.901e+02, threshold=5.537e+02, percent-clipped=1.0 +2023-02-09 00:32:15,428 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.2619, 5.2335, 4.7221, 2.5798, 4.6875, 4.9655, 4.8841, 4.7320], + device='cuda:1'), covar=tensor([0.0541, 0.0438, 0.0987, 0.4649, 0.0782, 0.0966, 0.1104, 0.0630], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0459, 0.0448, 0.0560, 0.0444, 0.0467, 0.0442, 0.0408], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:32:17,524 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222120.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:32:30,075 INFO [train.py:901] (1/4) Epoch 28, batch 3900, loss[loss=0.1971, simple_loss=0.2827, pruned_loss=0.0558, over 8478.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2837, pruned_loss=0.05779, over 1617853.70 frames. ], batch size: 29, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:33:00,581 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4504, 2.3132, 1.7460, 2.2364, 1.8939, 1.3792, 1.8596, 2.1575], + device='cuda:1'), covar=tensor([0.1520, 0.0513, 0.1417, 0.0622, 0.0939, 0.2045, 0.1221, 0.0844], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0245, 0.0343, 0.0315, 0.0305, 0.0348, 0.0350, 0.0323], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:33:06,447 INFO [train.py:901] (1/4) Epoch 28, batch 3950, loss[loss=0.2116, simple_loss=0.2981, pruned_loss=0.06254, over 8488.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2841, pruned_loss=0.05814, over 1617515.90 frames. ], batch size: 25, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:33:17,877 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222203.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:20,797 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.1081, 1.6695, 1.5357, 1.5300, 1.3756, 1.3786, 1.3705, 1.3013], + device='cuda:1'), covar=tensor([0.1196, 0.0565, 0.1343, 0.0661, 0.0940, 0.1570, 0.0962, 0.0897], + device='cuda:1'), in_proj_covar=tensor([0.0358, 0.0245, 0.0342, 0.0314, 0.0304, 0.0347, 0.0350, 0.0323], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:33:26,088 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.641e+02 2.338e+02 2.821e+02 3.606e+02 1.107e+03, threshold=5.643e+02, percent-clipped=4.0 +2023-02-09 00:33:31,794 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222223.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:35,381 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4065, 1.4052, 1.7628, 1.0889, 1.0275, 1.7759, 0.2735, 1.1721], + device='cuda:1'), covar=tensor([0.1465, 0.1087, 0.0420, 0.1033, 0.2355, 0.0402, 0.1727, 0.1249], + device='cuda:1'), in_proj_covar=tensor([0.0202, 0.0207, 0.0137, 0.0225, 0.0279, 0.0147, 0.0176, 0.0200], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 00:33:37,686 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-02-09 00:33:40,241 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222235.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:42,144 INFO [train.py:901] (1/4) Epoch 28, batch 4000, loss[loss=0.1709, simple_loss=0.2492, pruned_loss=0.04635, over 7930.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2827, pruned_loss=0.05738, over 1618893.55 frames. ], batch size: 20, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:33:43,719 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222240.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:51,744 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222252.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:34:01,378 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222265.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:34:17,514 INFO [train.py:901] (1/4) Epoch 28, batch 4050, loss[loss=0.2327, simple_loss=0.2996, pruned_loss=0.08289, over 8031.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2841, pruned_loss=0.05813, over 1625762.80 frames. ], batch size: 22, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:34:38,323 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.744e+02 2.414e+02 3.092e+02 4.009e+02 1.246e+03, threshold=6.184e+02, percent-clipped=7.0 +2023-02-09 00:34:54,250 INFO [train.py:901] (1/4) Epoch 28, batch 4100, loss[loss=0.1904, simple_loss=0.2814, pruned_loss=0.04971, over 8138.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2838, pruned_loss=0.05825, over 1622846.05 frames. ], batch size: 22, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:35:14,871 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222367.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:35:17,641 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222371.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:35:29,374 INFO [train.py:901] (1/4) Epoch 28, batch 4150, loss[loss=0.1685, simple_loss=0.2594, pruned_loss=0.03877, over 8297.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.284, pruned_loss=0.05824, over 1621943.95 frames. ], batch size: 23, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:35:49,101 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.589e+02 2.330e+02 2.692e+02 3.176e+02 6.436e+02, threshold=5.384e+02, percent-clipped=1.0 +2023-02-09 00:36:07,144 INFO [train.py:901] (1/4) Epoch 28, batch 4200, loss[loss=0.2717, simple_loss=0.3446, pruned_loss=0.09941, over 8475.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2839, pruned_loss=0.05821, over 1620269.24 frames. ], batch size: 49, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:36:13,556 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0066, 2.1649, 1.8311, 2.9064, 1.3259, 1.7382, 2.1631, 2.1450], + device='cuda:1'), covar=tensor([0.0776, 0.0871, 0.0879, 0.0349, 0.1189, 0.1269, 0.0818, 0.0891], + device='cuda:1'), in_proj_covar=tensor([0.0232, 0.0195, 0.0244, 0.0214, 0.0204, 0.0247, 0.0251, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 00:36:14,729 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-02-09 00:36:37,838 WARNING [train.py:1067] (1/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-02-09 00:36:41,343 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222486.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:36:42,592 INFO [train.py:901] (1/4) Epoch 28, batch 4250, loss[loss=0.1986, simple_loss=0.2894, pruned_loss=0.05388, over 8296.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2832, pruned_loss=0.05789, over 1617585.17 frames. ], batch size: 23, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:36:44,926 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222491.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:00,852 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.734e+02 2.539e+02 3.193e+02 4.198e+02 8.289e+02, threshold=6.386e+02, percent-clipped=5.0 +2023-02-09 00:37:01,758 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222516.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:18,018 INFO [train.py:901] (1/4) Epoch 28, batch 4300, loss[loss=0.2152, simple_loss=0.306, pruned_loss=0.06219, over 8342.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2836, pruned_loss=0.05821, over 1616226.42 frames. ], batch size: 26, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:37:20,160 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222541.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:25,011 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222547.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:30,690 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2308, 1.5549, 1.5441, 1.0325, 1.5668, 1.2369, 0.3241, 1.5137], + device='cuda:1'), covar=tensor([0.0813, 0.0488, 0.0455, 0.0744, 0.0632, 0.1197, 0.1225, 0.0408], + device='cuda:1'), in_proj_covar=tensor([0.0471, 0.0410, 0.0362, 0.0458, 0.0394, 0.0549, 0.0402, 0.0441], + device='cuda:1'), out_proj_covar=tensor([1.2467e-04, 1.0614e-04, 9.4104e-05, 1.1964e-04, 1.0311e-04, 1.5315e-04, + 1.0740e-04, 1.1552e-04], device='cuda:1') +2023-02-09 00:37:39,468 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222567.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:54,079 INFO [train.py:901] (1/4) Epoch 28, batch 4350, loss[loss=0.18, simple_loss=0.2718, pruned_loss=0.04412, over 8088.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2838, pruned_loss=0.05779, over 1617856.19 frames. ], batch size: 21, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:37:54,214 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222588.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:11,696 WARNING [train.py:1067] (1/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-02-09 00:38:13,114 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.579e+02 2.501e+02 2.979e+02 3.614e+02 7.360e+02, threshold=5.959e+02, percent-clipped=2.0 +2023-02-09 00:38:18,808 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222623.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:29,009 INFO [train.py:901] (1/4) Epoch 28, batch 4400, loss[loss=0.1796, simple_loss=0.2602, pruned_loss=0.04947, over 7662.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.283, pruned_loss=0.05759, over 1613455.15 frames. ], batch size: 19, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:38:32,613 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0626, 1.2874, 1.5934, 1.2129, 1.0463, 1.3656, 1.9803, 1.7456], + device='cuda:1'), covar=tensor([0.0554, 0.1713, 0.2338, 0.1890, 0.0672, 0.2074, 0.0737, 0.0656], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0190, 0.0162, 0.0101, 0.0163, 0.0113, 0.0146], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 00:38:36,656 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222648.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:47,075 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222662.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:47,604 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.2461, 4.2269, 3.8282, 1.9135, 3.7440, 3.8974, 3.7935, 3.6074], + device='cuda:1'), covar=tensor([0.0683, 0.0529, 0.1039, 0.4788, 0.0821, 0.0797, 0.1279, 0.0723], + device='cuda:1'), in_proj_covar=tensor([0.0544, 0.0458, 0.0450, 0.0559, 0.0444, 0.0464, 0.0440, 0.0407], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:38:54,364 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-02-09 00:39:01,989 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222682.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:39:05,978 INFO [train.py:901] (1/4) Epoch 28, batch 4450, loss[loss=0.1782, simple_loss=0.2638, pruned_loss=0.04634, over 7778.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.282, pruned_loss=0.0571, over 1611835.05 frames. ], batch size: 19, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:39:24,976 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.632e+02 2.353e+02 2.798e+02 3.446e+02 6.111e+02, threshold=5.597e+02, percent-clipped=1.0 +2023-02-09 00:39:29,406 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2121, 3.4334, 2.2462, 2.9502, 2.6937, 2.0638, 2.6917, 2.9465], + device='cuda:1'), covar=tensor([0.1616, 0.0456, 0.1236, 0.0653, 0.0837, 0.1502, 0.1063, 0.1112], + device='cuda:1'), in_proj_covar=tensor([0.0361, 0.0247, 0.0344, 0.0317, 0.0306, 0.0351, 0.0351, 0.0324], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:39:41,225 INFO [train.py:901] (1/4) Epoch 28, batch 4500, loss[loss=0.2085, simple_loss=0.2962, pruned_loss=0.06037, over 8110.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2818, pruned_loss=0.05689, over 1614388.29 frames. ], batch size: 23, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:39:42,067 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5547, 2.1519, 3.2068, 1.3924, 2.3990, 1.9895, 1.7621, 2.6317], + device='cuda:1'), covar=tensor([0.1968, 0.2563, 0.0802, 0.4750, 0.1950, 0.3355, 0.2507, 0.2111], + device='cuda:1'), in_proj_covar=tensor([0.0539, 0.0638, 0.0566, 0.0673, 0.0662, 0.0611, 0.0565, 0.0645], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:39:43,394 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.8535, 3.8357, 3.4241, 1.8662, 3.3643, 3.4798, 3.4535, 3.3598], + device='cuda:1'), covar=tensor([0.0869, 0.0698, 0.1188, 0.4639, 0.1020, 0.1182, 0.1404, 0.1010], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0457, 0.0449, 0.0557, 0.0442, 0.0463, 0.0439, 0.0406], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:39:44,228 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222742.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:39:45,367 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-02-09 00:40:02,731 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222767.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:40:18,345 INFO [train.py:901] (1/4) Epoch 28, batch 4550, loss[loss=0.1801, simple_loss=0.2697, pruned_loss=0.04524, over 8493.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2806, pruned_loss=0.05634, over 1609865.46 frames. ], batch size: 28, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:40:37,191 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.362e+02 2.324e+02 2.721e+02 3.677e+02 6.861e+02, threshold=5.442e+02, percent-clipped=4.0 +2023-02-09 00:40:53,695 INFO [train.py:901] (1/4) Epoch 28, batch 4600, loss[loss=0.2213, simple_loss=0.3049, pruned_loss=0.06886, over 8592.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2809, pruned_loss=0.05637, over 1613162.45 frames. ], batch size: 31, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:41:27,918 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222885.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:41:30,003 INFO [train.py:901] (1/4) Epoch 28, batch 4650, loss[loss=0.1896, simple_loss=0.2762, pruned_loss=0.05155, over 8240.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2812, pruned_loss=0.05674, over 1616170.14 frames. ], batch size: 24, lr: 2.68e-03, grad_scale: 16.0 +2023-02-09 00:41:30,921 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8038, 1.9692, 2.0730, 1.4910, 2.2161, 1.5406, 0.7883, 2.0164], + device='cuda:1'), covar=tensor([0.0657, 0.0417, 0.0380, 0.0701, 0.0519, 0.0983, 0.1033, 0.0364], + device='cuda:1'), in_proj_covar=tensor([0.0472, 0.0411, 0.0364, 0.0459, 0.0396, 0.0552, 0.0402, 0.0442], + device='cuda:1'), out_proj_covar=tensor([1.2487e-04, 1.0650e-04, 9.4666e-05, 1.2003e-04, 1.0351e-04, 1.5409e-04, + 1.0734e-04, 1.1564e-04], device='cuda:1') +2023-02-09 00:41:50,684 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.622e+02 2.423e+02 3.099e+02 3.500e+02 7.849e+02, threshold=6.198e+02, percent-clipped=6.0 +2023-02-09 00:41:53,071 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222918.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:41:55,969 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.8962, 2.4200, 3.7817, 2.0278, 2.1350, 3.7456, 0.6897, 2.3652], + device='cuda:1'), covar=tensor([0.1412, 0.1057, 0.0187, 0.1351, 0.2060, 0.0210, 0.2027, 0.1081], + device='cuda:1'), in_proj_covar=tensor([0.0202, 0.0206, 0.0138, 0.0225, 0.0279, 0.0148, 0.0176, 0.0200], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 00:42:02,696 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222932.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:06,748 INFO [train.py:901] (1/4) Epoch 28, batch 4700, loss[loss=0.1945, simple_loss=0.2699, pruned_loss=0.05953, over 7533.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2802, pruned_loss=0.05624, over 1614505.40 frames. ], batch size: 18, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:42:06,964 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222938.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:10,432 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222943.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:20,638 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222958.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:24,246 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222963.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:40,837 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222987.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:41,329 INFO [train.py:901] (1/4) Epoch 28, batch 4750, loss[loss=0.1795, simple_loss=0.2641, pruned_loss=0.04748, over 7803.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2803, pruned_loss=0.05638, over 1615903.05 frames. ], batch size: 20, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:42:50,010 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223000.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:52,055 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223002.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:54,708 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-02-09 00:42:58,161 WARNING [train.py:1067] (1/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-02-09 00:43:02,879 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.833e+02 2.412e+02 2.807e+02 3.833e+02 7.869e+02, threshold=5.613e+02, percent-clipped=5.0 +2023-02-09 00:43:18,663 INFO [train.py:901] (1/4) Epoch 28, batch 4800, loss[loss=0.1847, simple_loss=0.2683, pruned_loss=0.05051, over 8294.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.281, pruned_loss=0.05668, over 1620181.79 frames. ], batch size: 23, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:43:25,185 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223047.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:43:32,780 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3418, 2.0858, 1.6313, 1.9802, 1.7038, 1.4370, 1.6787, 1.7668], + device='cuda:1'), covar=tensor([0.1411, 0.0494, 0.1360, 0.0559, 0.0801, 0.1609, 0.0953, 0.0897], + device='cuda:1'), in_proj_covar=tensor([0.0358, 0.0244, 0.0340, 0.0313, 0.0302, 0.0346, 0.0347, 0.0321], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:43:47,828 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.46 vs. limit=5.0 +2023-02-09 00:43:48,814 WARNING [train.py:1067] (1/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-02-09 00:43:49,658 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223082.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:43:53,653 INFO [train.py:901] (1/4) Epoch 28, batch 4850, loss[loss=0.2073, simple_loss=0.2989, pruned_loss=0.05785, over 8496.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.281, pruned_loss=0.05659, over 1618410.76 frames. ], batch size: 26, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:44:10,873 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-02-09 00:44:13,745 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.627e+02 2.508e+02 3.332e+02 4.408e+02 9.671e+02, threshold=6.663e+02, percent-clipped=7.0 +2023-02-09 00:44:31,160 INFO [train.py:901] (1/4) Epoch 28, batch 4900, loss[loss=0.2047, simple_loss=0.2874, pruned_loss=0.06105, over 8508.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2817, pruned_loss=0.05718, over 1615253.50 frames. ], batch size: 26, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:44:48,259 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5654, 1.3283, 2.6775, 1.0628, 2.1553, 2.8374, 3.1801, 2.0839], + device='cuda:1'), covar=tensor([0.1575, 0.2205, 0.0542, 0.3099, 0.1098, 0.0486, 0.0653, 0.1213], + device='cuda:1'), in_proj_covar=tensor([0.0305, 0.0326, 0.0292, 0.0320, 0.0322, 0.0275, 0.0439, 0.0308], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 00:44:48,925 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.1287, 1.4870, 4.3781, 1.6174, 3.8831, 3.6280, 3.9623, 3.8861], + device='cuda:1'), covar=tensor([0.0695, 0.4689, 0.0590, 0.4119, 0.1118, 0.1029, 0.0631, 0.0671], + device='cuda:1'), in_proj_covar=tensor([0.0689, 0.0674, 0.0743, 0.0667, 0.0750, 0.0643, 0.0649, 0.0722], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:45:07,006 INFO [train.py:901] (1/4) Epoch 28, batch 4950, loss[loss=0.1787, simple_loss=0.259, pruned_loss=0.04917, over 7799.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2818, pruned_loss=0.05686, over 1615840.12 frames. ], batch size: 19, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:45:09,273 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223191.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:45:22,625 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1861, 2.5715, 3.6051, 2.0958, 3.0445, 2.5871, 2.3162, 2.9306], + device='cuda:1'), covar=tensor([0.1550, 0.2162, 0.0812, 0.3661, 0.1412, 0.2527, 0.1986, 0.2107], + device='cuda:1'), in_proj_covar=tensor([0.0541, 0.0638, 0.0565, 0.0673, 0.0664, 0.0612, 0.0565, 0.0648], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:45:26,420 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.723e+02 2.334e+02 2.712e+02 3.560e+02 9.309e+02, threshold=5.424e+02, percent-clipped=3.0 +2023-02-09 00:45:42,295 INFO [train.py:901] (1/4) Epoch 28, batch 5000, loss[loss=0.1827, simple_loss=0.2574, pruned_loss=0.05401, over 7424.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2822, pruned_loss=0.05704, over 1614837.59 frames. ], batch size: 17, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:45:56,134 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223256.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:14,345 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223281.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:18,935 INFO [train.py:901] (1/4) Epoch 28, batch 5050, loss[loss=0.1777, simple_loss=0.2701, pruned_loss=0.04264, over 8101.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2824, pruned_loss=0.05724, over 1617506.70 frames. ], batch size: 23, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:46:28,802 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223302.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:29,633 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223303.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:32,889 WARNING [train.py:1067] (1/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-02-09 00:46:38,387 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.592e+02 2.274e+02 2.931e+02 3.573e+02 6.090e+02, threshold=5.862e+02, percent-clipped=1.0 +2023-02-09 00:46:46,948 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223328.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:48,965 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223331.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:53,555 INFO [train.py:901] (1/4) Epoch 28, batch 5100, loss[loss=0.1947, simple_loss=0.2828, pruned_loss=0.05329, over 8027.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2813, pruned_loss=0.0564, over 1617448.15 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:46:59,471 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223346.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:47:26,338 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2672, 1.6976, 4.3074, 2.1577, 2.5365, 4.8962, 5.0360, 4.2654], + device='cuda:1'), covar=tensor([0.1180, 0.1882, 0.0271, 0.1904, 0.1216, 0.0182, 0.0408, 0.0534], + device='cuda:1'), in_proj_covar=tensor([0.0305, 0.0325, 0.0292, 0.0319, 0.0322, 0.0274, 0.0439, 0.0307], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 00:47:31,192 INFO [train.py:901] (1/4) Epoch 28, batch 5150, loss[loss=0.2071, simple_loss=0.2875, pruned_loss=0.06329, over 8576.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2806, pruned_loss=0.05629, over 1616708.48 frames. ], batch size: 49, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:47:37,584 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5113, 1.7809, 1.8343, 1.2664, 1.8924, 1.4045, 0.5099, 1.7596], + device='cuda:1'), covar=tensor([0.0634, 0.0426, 0.0355, 0.0630, 0.0417, 0.1035, 0.1004, 0.0324], + device='cuda:1'), in_proj_covar=tensor([0.0469, 0.0409, 0.0362, 0.0456, 0.0394, 0.0549, 0.0400, 0.0438], + device='cuda:1'), out_proj_covar=tensor([1.2407e-04, 1.0587e-04, 9.4242e-05, 1.1912e-04, 1.0306e-04, 1.5314e-04, + 1.0684e-04, 1.1481e-04], device='cuda:1') +2023-02-09 00:47:38,172 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.6105, 4.6190, 4.1862, 2.3797, 4.1176, 4.1828, 4.1793, 4.0159], + device='cuda:1'), covar=tensor([0.0733, 0.0520, 0.0954, 0.3824, 0.0837, 0.0934, 0.1261, 0.0642], + device='cuda:1'), in_proj_covar=tensor([0.0539, 0.0454, 0.0449, 0.0554, 0.0443, 0.0463, 0.0441, 0.0407], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:47:43,144 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.31 vs. limit=5.0 +2023-02-09 00:47:50,489 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.755e+02 2.349e+02 2.964e+02 3.516e+02 1.122e+03, threshold=5.928e+02, percent-clipped=3.0 +2023-02-09 00:47:51,397 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223417.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:47:57,693 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223426.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:47:58,560 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4589, 1.3858, 1.7919, 1.2243, 1.0445, 1.7805, 0.2501, 1.1598], + device='cuda:1'), covar=tensor([0.1365, 0.1238, 0.0422, 0.0854, 0.2574, 0.0442, 0.1905, 0.1140], + device='cuda:1'), in_proj_covar=tensor([0.0200, 0.0206, 0.0137, 0.0224, 0.0279, 0.0148, 0.0175, 0.0200], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 00:48:05,964 INFO [train.py:901] (1/4) Epoch 28, batch 5200, loss[loss=0.188, simple_loss=0.2602, pruned_loss=0.05793, over 6839.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2817, pruned_loss=0.05697, over 1618836.30 frames. ], batch size: 15, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:48:11,623 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223446.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:48:22,805 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223461.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:48:31,242 WARNING [train.py:1067] (1/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-02-09 00:48:44,075 INFO [train.py:901] (1/4) Epoch 28, batch 5250, loss[loss=0.2114, simple_loss=0.2922, pruned_loss=0.06533, over 7695.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2814, pruned_loss=0.05699, over 1613139.79 frames. ], batch size: 18, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:49:03,771 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.671e+02 2.237e+02 2.837e+02 3.561e+02 7.405e+02, threshold=5.674e+02, percent-clipped=6.0 +2023-02-09 00:49:17,138 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223535.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:49:19,101 INFO [train.py:901] (1/4) Epoch 28, batch 5300, loss[loss=0.2256, simple_loss=0.3078, pruned_loss=0.07171, over 8356.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2813, pruned_loss=0.0567, over 1613357.80 frames. ], batch size: 24, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:49:21,372 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223541.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:49:22,687 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1115, 2.0258, 2.5869, 2.2100, 2.5727, 2.2362, 2.0445, 1.4844], + device='cuda:1'), covar=tensor([0.5787, 0.4884, 0.1987, 0.3830, 0.2456, 0.3138, 0.1966, 0.5397], + device='cuda:1'), in_proj_covar=tensor([0.0966, 0.1030, 0.0835, 0.0999, 0.1026, 0.0936, 0.0773, 0.0855], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 00:49:55,542 INFO [train.py:901] (1/4) Epoch 28, batch 5350, loss[loss=0.2405, simple_loss=0.3208, pruned_loss=0.08011, over 7245.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2826, pruned_loss=0.05788, over 1608285.55 frames. ], batch size: 73, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:50:01,803 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223596.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:50:04,677 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.8428, 1.5018, 4.0046, 1.4328, 3.5620, 3.3245, 3.6187, 3.5409], + device='cuda:1'), covar=tensor([0.0738, 0.4515, 0.0608, 0.4482, 0.1184, 0.1063, 0.0743, 0.0771], + device='cuda:1'), in_proj_covar=tensor([0.0679, 0.0663, 0.0734, 0.0656, 0.0737, 0.0632, 0.0640, 0.0712], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:50:15,659 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.359e+02 2.840e+02 3.657e+02 7.209e+02, threshold=5.681e+02, percent-clipped=3.0 +2023-02-09 00:50:30,412 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2841, 2.0376, 2.5978, 2.2233, 2.5509, 2.3488, 2.2000, 1.4277], + device='cuda:1'), covar=tensor([0.5604, 0.4992, 0.2158, 0.3906, 0.2600, 0.3240, 0.1931, 0.5664], + device='cuda:1'), in_proj_covar=tensor([0.0967, 0.1031, 0.0837, 0.1000, 0.1027, 0.0938, 0.0774, 0.0856], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 00:50:30,870 INFO [train.py:901] (1/4) Epoch 28, batch 5400, loss[loss=0.1927, simple_loss=0.2651, pruned_loss=0.06012, over 8078.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2833, pruned_loss=0.0582, over 1611779.05 frames. ], batch size: 21, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:50:39,702 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223650.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:50:55,924 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223673.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:03,756 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2919, 3.3097, 2.3535, 2.7445, 2.5293, 2.1670, 2.5896, 2.9725], + device='cuda:1'), covar=tensor([0.1353, 0.0339, 0.1052, 0.0714, 0.0758, 0.1417, 0.0958, 0.0922], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0244, 0.0342, 0.0315, 0.0303, 0.0347, 0.0350, 0.0323], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:51:06,360 INFO [train.py:901] (1/4) Epoch 28, batch 5450, loss[loss=0.1785, simple_loss=0.2548, pruned_loss=0.05109, over 7552.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2833, pruned_loss=0.05804, over 1609888.13 frames. ], batch size: 18, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:51:13,800 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223698.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:16,752 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223702.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:28,428 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.718e+02 2.405e+02 2.886e+02 3.694e+02 6.837e+02, threshold=5.773e+02, percent-clipped=3.0 +2023-02-09 00:51:28,657 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223717.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:29,453 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2626, 2.0976, 2.7042, 2.3057, 2.7474, 2.3579, 2.1487, 1.5821], + device='cuda:1'), covar=tensor([0.5747, 0.5260, 0.2281, 0.3831, 0.2531, 0.3336, 0.1941, 0.5558], + device='cuda:1'), in_proj_covar=tensor([0.0966, 0.1029, 0.0837, 0.0998, 0.1025, 0.0934, 0.0772, 0.0855], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 00:51:31,414 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-02-09 00:51:35,317 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-02-09 00:51:36,523 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223727.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:44,560 INFO [train.py:901] (1/4) Epoch 28, batch 5500, loss[loss=0.2057, simple_loss=0.2932, pruned_loss=0.05905, over 8249.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.282, pruned_loss=0.05769, over 1607409.53 frames. ], batch size: 24, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:51:47,326 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223742.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:52:18,971 INFO [train.py:901] (1/4) Epoch 28, batch 5550, loss[loss=0.1621, simple_loss=0.2413, pruned_loss=0.04145, over 7442.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2829, pruned_loss=0.05831, over 1610939.24 frames. ], batch size: 17, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:52:20,246 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-02-09 00:52:25,540 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223797.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:52:39,665 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.770e+02 2.462e+02 3.010e+02 3.574e+02 1.274e+03, threshold=6.020e+02, percent-clipped=3.0 +2023-02-09 00:52:43,459 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223822.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:52:55,577 INFO [train.py:901] (1/4) Epoch 28, batch 5600, loss[loss=0.1921, simple_loss=0.2725, pruned_loss=0.05581, over 7930.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2818, pruned_loss=0.05767, over 1609302.20 frames. ], batch size: 20, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:53:05,100 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5479, 1.4594, 1.8226, 1.2895, 1.2063, 1.8376, 0.2561, 1.2281], + device='cuda:1'), covar=tensor([0.1384, 0.1256, 0.0408, 0.0725, 0.2252, 0.0440, 0.1819, 0.1098], + device='cuda:1'), in_proj_covar=tensor([0.0201, 0.0205, 0.0137, 0.0223, 0.0278, 0.0147, 0.0174, 0.0199], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 00:53:31,691 INFO [train.py:901] (1/4) Epoch 28, batch 5650, loss[loss=0.1885, simple_loss=0.2805, pruned_loss=0.04828, over 8763.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2817, pruned_loss=0.05785, over 1610209.30 frames. ], batch size: 30, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:53:41,197 WARNING [train.py:1067] (1/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-02-09 00:53:44,191 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223906.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:53:44,786 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7421, 1.8255, 1.5942, 2.3379, 1.0076, 1.4730, 1.6512, 1.7496], + device='cuda:1'), covar=tensor([0.0769, 0.0696, 0.0870, 0.0388, 0.1039, 0.1253, 0.0744, 0.0755], + device='cuda:1'), in_proj_covar=tensor([0.0233, 0.0196, 0.0246, 0.0215, 0.0205, 0.0249, 0.0252, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 00:53:51,298 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.465e+02 2.313e+02 2.789e+02 3.752e+02 1.102e+03, threshold=5.578e+02, percent-clipped=3.0 +2023-02-09 00:54:01,602 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223931.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 00:54:06,155 INFO [train.py:901] (1/4) Epoch 28, batch 5700, loss[loss=0.2186, simple_loss=0.3107, pruned_loss=0.06322, over 8500.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.282, pruned_loss=0.05777, over 1613612.23 frames. ], batch size: 28, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:54:07,519 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223940.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:54:10,615 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-02-09 00:54:16,420 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-02-09 00:54:43,234 INFO [train.py:901] (1/4) Epoch 28, batch 5750, loss[loss=0.2151, simple_loss=0.2935, pruned_loss=0.06838, over 7660.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2817, pruned_loss=0.05749, over 1612065.23 frames. ], batch size: 19, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:54:48,708 WARNING [train.py:1067] (1/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-02-09 00:55:04,150 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.665e+02 2.263e+02 2.713e+02 3.241e+02 8.661e+02, threshold=5.425e+02, percent-clipped=3.0 +2023-02-09 00:55:18,773 INFO [train.py:901] (1/4) Epoch 28, batch 5800, loss[loss=0.1879, simple_loss=0.2743, pruned_loss=0.05073, over 8615.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2812, pruned_loss=0.05684, over 1611567.70 frames. ], batch size: 39, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:55:22,317 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.7124, 1.6448, 5.9280, 2.2830, 5.3101, 5.0035, 5.4546, 5.3643], + device='cuda:1'), covar=tensor([0.0551, 0.5123, 0.0422, 0.4025, 0.1039, 0.0950, 0.0497, 0.0566], + device='cuda:1'), in_proj_covar=tensor([0.0683, 0.0665, 0.0737, 0.0661, 0.0742, 0.0636, 0.0643, 0.0718], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:55:30,597 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=224055.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:55:31,242 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.2370, 3.1652, 2.9191, 1.6450, 2.8845, 2.8981, 2.7830, 2.8658], + device='cuda:1'), covar=tensor([0.1122, 0.0783, 0.1252, 0.4489, 0.1167, 0.1246, 0.1616, 0.1052], + device='cuda:1'), in_proj_covar=tensor([0.0546, 0.0458, 0.0452, 0.0560, 0.0446, 0.0467, 0.0444, 0.0410], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:55:43,054 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7045, 2.1086, 3.2581, 1.5313, 2.4701, 2.1368, 1.8436, 2.5085], + device='cuda:1'), covar=tensor([0.1941, 0.2713, 0.0815, 0.4771, 0.1879, 0.3222, 0.2458, 0.2345], + device='cuda:1'), in_proj_covar=tensor([0.0541, 0.0636, 0.0564, 0.0671, 0.0663, 0.0612, 0.0563, 0.0647], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:55:47,081 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4811, 2.3167, 1.7653, 2.1434, 1.9236, 1.4495, 1.8958, 1.9611], + device='cuda:1'), covar=tensor([0.1569, 0.0478, 0.1390, 0.0690, 0.0814, 0.1727, 0.1015, 0.1027], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0243, 0.0341, 0.0314, 0.0303, 0.0346, 0.0350, 0.0322], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 00:55:55,788 INFO [train.py:901] (1/4) Epoch 28, batch 5850, loss[loss=0.2104, simple_loss=0.2849, pruned_loss=0.06794, over 8295.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2814, pruned_loss=0.05649, over 1613458.04 frames. ], batch size: 23, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:56:15,667 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.578e+02 2.538e+02 3.148e+02 4.118e+02 7.183e+02, threshold=6.296e+02, percent-clipped=12.0 +2023-02-09 00:56:30,204 INFO [train.py:901] (1/4) Epoch 28, batch 5900, loss[loss=0.1821, simple_loss=0.2528, pruned_loss=0.05577, over 7779.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2817, pruned_loss=0.05717, over 1613728.54 frames. ], batch size: 19, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:57:06,180 INFO [train.py:901] (1/4) Epoch 28, batch 5950, loss[loss=0.1811, simple_loss=0.2737, pruned_loss=0.04423, over 8240.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.281, pruned_loss=0.05703, over 1611639.03 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:57:25,370 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-02-09 00:57:28,301 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.818e+02 2.486e+02 3.110e+02 3.888e+02 7.674e+02, threshold=6.220e+02, percent-clipped=4.0 +2023-02-09 00:57:37,734 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224230.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:57:43,109 INFO [train.py:901] (1/4) Epoch 28, batch 6000, loss[loss=0.1825, simple_loss=0.2637, pruned_loss=0.05065, over 7940.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2808, pruned_loss=0.05669, over 1611932.37 frames. ], batch size: 20, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:57:43,110 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 00:57:56,800 INFO [train.py:935] (1/4) Epoch 28, validation: loss=0.1714, simple_loss=0.2708, pruned_loss=0.03603, over 944034.00 frames. +2023-02-09 00:57:56,801 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 00:57:59,119 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0119, 1.8217, 2.2815, 1.9509, 2.2758, 2.1002, 1.9133, 1.2537], + device='cuda:1'), covar=tensor([0.6243, 0.5460, 0.2100, 0.3952, 0.2582, 0.3489, 0.2087, 0.5456], + device='cuda:1'), in_proj_covar=tensor([0.0973, 0.1035, 0.0841, 0.1004, 0.1028, 0.0940, 0.0776, 0.0859], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 00:58:07,094 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.7658, 5.9471, 5.2271, 2.3518, 5.2449, 5.5306, 5.3643, 5.3950], + device='cuda:1'), covar=tensor([0.0455, 0.0297, 0.0725, 0.4238, 0.0659, 0.0680, 0.0897, 0.0565], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0457, 0.0451, 0.0559, 0.0445, 0.0466, 0.0442, 0.0408], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:58:33,310 INFO [train.py:901] (1/4) Epoch 28, batch 6050, loss[loss=0.1988, simple_loss=0.2858, pruned_loss=0.05594, over 8644.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2807, pruned_loss=0.05628, over 1613146.54 frames. ], batch size: 34, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:58:43,545 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6209, 2.0269, 3.2759, 1.5017, 2.4673, 2.1456, 1.7335, 2.5762], + device='cuda:1'), covar=tensor([0.1884, 0.2800, 0.0787, 0.4623, 0.1895, 0.3155, 0.2438, 0.2256], + device='cuda:1'), in_proj_covar=tensor([0.0541, 0.0637, 0.0565, 0.0671, 0.0665, 0.0614, 0.0565, 0.0649], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 00:58:48,812 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-02-09 00:58:49,932 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=224311.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:58:52,994 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-02-09 00:58:54,040 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.623e+02 2.462e+02 3.109e+02 3.867e+02 1.260e+03, threshold=6.217e+02, percent-clipped=5.0 +2023-02-09 00:59:08,857 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=224336.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:59:10,069 INFO [train.py:901] (1/4) Epoch 28, batch 6100, loss[loss=0.194, simple_loss=0.2689, pruned_loss=0.05953, over 7812.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2818, pruned_loss=0.05682, over 1612663.13 frames. ], batch size: 20, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:59:26,305 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-02-09 00:59:46,740 INFO [train.py:901] (1/4) Epoch 28, batch 6150, loss[loss=0.1994, simple_loss=0.2891, pruned_loss=0.05488, over 8354.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.281, pruned_loss=0.05641, over 1612053.10 frames. ], batch size: 24, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:59:54,603 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0684, 2.2492, 2.2886, 1.5890, 2.3963, 1.8230, 1.7306, 2.0423], + device='cuda:1'), covar=tensor([0.0675, 0.0430, 0.0344, 0.0705, 0.0485, 0.0716, 0.0780, 0.0465], + device='cuda:1'), in_proj_covar=tensor([0.0471, 0.0409, 0.0363, 0.0458, 0.0393, 0.0550, 0.0402, 0.0440], + device='cuda:1'), out_proj_covar=tensor([1.2457e-04, 1.0580e-04, 9.4517e-05, 1.1960e-04, 1.0283e-04, 1.5344e-04, + 1.0723e-04, 1.1521e-04], device='cuda:1') +2023-02-09 01:00:06,960 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.535e+02 2.362e+02 2.823e+02 3.455e+02 8.158e+02, threshold=5.645e+02, percent-clipped=2.0 +2023-02-09 01:00:21,354 INFO [train.py:901] (1/4) Epoch 28, batch 6200, loss[loss=0.1771, simple_loss=0.2664, pruned_loss=0.04387, over 8088.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2818, pruned_loss=0.05648, over 1616770.29 frames. ], batch size: 21, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 01:00:58,233 INFO [train.py:901] (1/4) Epoch 28, batch 6250, loss[loss=0.1814, simple_loss=0.2563, pruned_loss=0.05326, over 7527.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2823, pruned_loss=0.05691, over 1618630.33 frames. ], batch size: 18, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 01:00:59,402 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.45 vs. limit=5.0 +2023-02-09 01:01:18,580 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.516e+02 2.593e+02 3.043e+02 4.250e+02 9.084e+02, threshold=6.087e+02, percent-clipped=11.0 +2023-02-09 01:01:33,267 INFO [train.py:901] (1/4) Epoch 28, batch 6300, loss[loss=0.1907, simple_loss=0.282, pruned_loss=0.04971, over 8286.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2815, pruned_loss=0.05636, over 1616286.82 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:02:00,128 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=224574.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:02:04,359 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224580.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:02:10,460 INFO [train.py:901] (1/4) Epoch 28, batch 6350, loss[loss=0.2133, simple_loss=0.2951, pruned_loss=0.06572, over 8118.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2806, pruned_loss=0.05617, over 1609053.31 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:02:30,928 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.559e+02 2.315e+02 2.720e+02 3.259e+02 6.733e+02, threshold=5.440e+02, percent-clipped=2.0 +2023-02-09 01:02:45,883 INFO [train.py:901] (1/4) Epoch 28, batch 6400, loss[loss=0.2312, simple_loss=0.3186, pruned_loss=0.07189, over 8461.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2806, pruned_loss=0.0563, over 1608084.70 frames. ], batch size: 27, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:03:16,424 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-09 01:03:21,510 INFO [train.py:901] (1/4) Epoch 28, batch 6450, loss[loss=0.1982, simple_loss=0.2902, pruned_loss=0.05306, over 8299.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2798, pruned_loss=0.05606, over 1604204.40 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:03:22,396 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=224689.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:03:43,002 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.578e+02 2.303e+02 2.784e+02 3.485e+02 7.082e+02, threshold=5.567e+02, percent-clipped=7.0 +2023-02-09 01:03:44,831 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-02-09 01:03:56,389 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2256, 2.4539, 2.7808, 1.7958, 3.0627, 1.8836, 1.4732, 2.2746], + device='cuda:1'), covar=tensor([0.0979, 0.0547, 0.0353, 0.0933, 0.0615, 0.0996, 0.1236, 0.0662], + device='cuda:1'), in_proj_covar=tensor([0.0475, 0.0412, 0.0366, 0.0461, 0.0396, 0.0555, 0.0406, 0.0444], + device='cuda:1'), out_proj_covar=tensor([1.2572e-04, 1.0680e-04, 9.5128e-05, 1.2062e-04, 1.0360e-04, 1.5473e-04, + 1.0836e-04, 1.1623e-04], device='cuda:1') +2023-02-09 01:03:57,597 INFO [train.py:901] (1/4) Epoch 28, batch 6500, loss[loss=0.2125, simple_loss=0.3005, pruned_loss=0.06227, over 8539.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2795, pruned_loss=0.05579, over 1605898.32 frames. ], batch size: 31, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:04:02,104 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224744.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:04:11,961 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224758.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:04:32,262 INFO [train.py:901] (1/4) Epoch 28, batch 6550, loss[loss=0.2184, simple_loss=0.2984, pruned_loss=0.06919, over 8495.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2806, pruned_loss=0.05632, over 1608716.32 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:04:47,798 WARNING [train.py:1067] (1/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-02-09 01:04:53,995 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.746e+02 2.489e+02 3.184e+02 3.768e+02 7.222e+02, threshold=6.368e+02, percent-clipped=1.0 +2023-02-09 01:05:08,029 WARNING [train.py:1067] (1/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-02-09 01:05:09,348 INFO [train.py:901] (1/4) Epoch 28, batch 6600, loss[loss=0.1895, simple_loss=0.2749, pruned_loss=0.05201, over 8028.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2804, pruned_loss=0.05602, over 1609609.67 frames. ], batch size: 22, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:05:17,179 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4246, 1.3394, 1.6728, 1.0999, 1.0877, 1.6595, 0.2891, 1.1237], + device='cuda:1'), covar=tensor([0.1427, 0.1124, 0.0396, 0.0887, 0.2167, 0.0449, 0.1815, 0.1241], + device='cuda:1'), in_proj_covar=tensor([0.0204, 0.0209, 0.0138, 0.0226, 0.0280, 0.0149, 0.0176, 0.0201], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 01:05:44,396 INFO [train.py:901] (1/4) Epoch 28, batch 6650, loss[loss=0.1734, simple_loss=0.2683, pruned_loss=0.0392, over 8186.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2816, pruned_loss=0.0568, over 1614147.38 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:06:04,780 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.705e+02 2.463e+02 2.971e+02 3.895e+02 9.422e+02, threshold=5.941e+02, percent-clipped=4.0 +2023-02-09 01:06:10,964 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=224924.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:06:11,638 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.7797, 5.9266, 5.1626, 2.7804, 5.2192, 5.6425, 5.4208, 5.3638], + device='cuda:1'), covar=tensor([0.0495, 0.0357, 0.0846, 0.3890, 0.0774, 0.0755, 0.1030, 0.0656], + device='cuda:1'), in_proj_covar=tensor([0.0550, 0.0462, 0.0456, 0.0566, 0.0448, 0.0471, 0.0448, 0.0414], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:06:17,461 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-02-09 01:06:21,167 INFO [train.py:901] (1/4) Epoch 28, batch 6700, loss[loss=0.1709, simple_loss=0.2534, pruned_loss=0.04418, over 5562.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2826, pruned_loss=0.05753, over 1613511.49 frames. ], batch size: 12, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:06:26,239 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=224945.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:06:30,484 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.1341, 1.3725, 1.5987, 1.3455, 0.8524, 1.4105, 1.2273, 1.0411], + device='cuda:1'), covar=tensor([0.0633, 0.1240, 0.1603, 0.1444, 0.0548, 0.1501, 0.0714, 0.0727], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0154, 0.0190, 0.0161, 0.0102, 0.0164, 0.0113, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:06:44,443 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=224970.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:06:57,024 INFO [train.py:901] (1/4) Epoch 28, batch 6750, loss[loss=0.1634, simple_loss=0.2402, pruned_loss=0.04329, over 7711.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2821, pruned_loss=0.05728, over 1614080.48 frames. ], batch size: 18, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:06:57,469 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.63 vs. limit=2.0 +2023-02-09 01:06:58,659 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8177, 2.5921, 1.9183, 2.4556, 2.3775, 1.6330, 2.3214, 2.3318], + device='cuda:1'), covar=tensor([0.1534, 0.0446, 0.1280, 0.0694, 0.0783, 0.1700, 0.1022, 0.1120], + device='cuda:1'), in_proj_covar=tensor([0.0362, 0.0247, 0.0345, 0.0317, 0.0304, 0.0350, 0.0354, 0.0327], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 01:07:01,382 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4169, 2.7419, 3.0060, 1.8004, 3.3448, 2.0369, 1.5247, 2.4366], + device='cuda:1'), covar=tensor([0.1008, 0.0434, 0.0419, 0.0928, 0.0463, 0.1084, 0.1160, 0.0644], + device='cuda:1'), in_proj_covar=tensor([0.0473, 0.0409, 0.0364, 0.0459, 0.0396, 0.0552, 0.0403, 0.0442], + device='cuda:1'), out_proj_covar=tensor([1.2520e-04, 1.0596e-04, 9.4811e-05, 1.2003e-04, 1.0343e-04, 1.5400e-04, + 1.0766e-04, 1.1576e-04], device='cuda:1') +2023-02-09 01:07:17,010 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.474e+02 2.343e+02 2.979e+02 3.883e+02 6.136e+02, threshold=5.958e+02, percent-clipped=2.0 +2023-02-09 01:07:28,015 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-02-09 01:07:32,093 INFO [train.py:901] (1/4) Epoch 28, batch 6800, loss[loss=0.2155, simple_loss=0.2994, pruned_loss=0.06583, over 8592.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2817, pruned_loss=0.05729, over 1615736.71 frames. ], batch size: 39, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:07:32,951 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225039.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:07:57,277 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.5590, 1.7740, 2.0839, 1.7803, 1.2093, 1.8710, 2.3378, 1.9050], + device='cuda:1'), covar=tensor([0.0549, 0.1185, 0.1525, 0.1400, 0.0627, 0.1368, 0.0651, 0.0665], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0161, 0.0102, 0.0163, 0.0113, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:08:08,382 INFO [train.py:901] (1/4) Epoch 28, batch 6850, loss[loss=0.185, simple_loss=0.2594, pruned_loss=0.05526, over 7799.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2821, pruned_loss=0.0571, over 1618913.25 frames. ], batch size: 19, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:08:08,454 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225088.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:08:18,039 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225102.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:08:18,737 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-02-09 01:08:28,500 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.612e+02 2.283e+02 2.996e+02 3.907e+02 8.918e+02, threshold=5.992e+02, percent-clipped=3.0 +2023-02-09 01:08:31,371 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225121.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:08:42,882 INFO [train.py:901] (1/4) Epoch 28, batch 6900, loss[loss=0.1636, simple_loss=0.2531, pruned_loss=0.03704, over 5096.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2831, pruned_loss=0.05735, over 1617152.45 frames. ], batch size: 11, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:08:47,492 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.66 vs. limit=2.0 +2023-02-09 01:08:58,668 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225160.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:09:18,975 INFO [train.py:901] (1/4) Epoch 28, batch 6950, loss[loss=0.1647, simple_loss=0.2527, pruned_loss=0.03838, over 8075.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2851, pruned_loss=0.05827, over 1620000.87 frames. ], batch size: 21, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:09:30,328 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225203.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:09:30,860 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-02-09 01:09:40,023 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.583e+02 2.456e+02 2.946e+02 3.977e+02 8.721e+02, threshold=5.892e+02, percent-clipped=6.0 +2023-02-09 01:09:40,234 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225217.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:09:54,781 INFO [train.py:901] (1/4) Epoch 28, batch 7000, loss[loss=0.1804, simple_loss=0.261, pruned_loss=0.04993, over 7276.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2836, pruned_loss=0.0577, over 1620087.09 frames. ], batch size: 16, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:10:31,216 INFO [train.py:901] (1/4) Epoch 28, batch 7050, loss[loss=0.2251, simple_loss=0.3025, pruned_loss=0.0738, over 8502.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2828, pruned_loss=0.05769, over 1618061.93 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:10:36,306 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225295.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:10:41,258 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.9223, 6.2328, 5.3009, 2.9806, 5.4099, 5.7729, 5.5764, 5.6207], + device='cuda:1'), covar=tensor([0.0557, 0.0343, 0.0922, 0.3670, 0.0780, 0.0998, 0.0913, 0.0623], + device='cuda:1'), in_proj_covar=tensor([0.0549, 0.0463, 0.0457, 0.0564, 0.0448, 0.0470, 0.0447, 0.0415], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:10:52,623 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.739e+02 2.402e+02 2.844e+02 3.449e+02 6.425e+02, threshold=5.688e+02, percent-clipped=2.0 +2023-02-09 01:10:55,647 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225320.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:11:08,201 INFO [train.py:901] (1/4) Epoch 28, batch 7100, loss[loss=0.1772, simple_loss=0.2673, pruned_loss=0.0435, over 8135.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.281, pruned_loss=0.05722, over 1617486.22 frames. ], batch size: 22, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:11:43,048 INFO [train.py:901] (1/4) Epoch 28, batch 7150, loss[loss=0.218, simple_loss=0.2883, pruned_loss=0.0738, over 8495.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2816, pruned_loss=0.05749, over 1611046.62 frames. ], batch size: 29, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:12:05,408 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.501e+02 2.377e+02 2.906e+02 3.542e+02 6.036e+02, threshold=5.811e+02, percent-clipped=2.0 +2023-02-09 01:12:21,600 INFO [train.py:901] (1/4) Epoch 28, batch 7200, loss[loss=0.1876, simple_loss=0.2738, pruned_loss=0.05068, over 8497.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2836, pruned_loss=0.0585, over 1614532.05 frames. ], batch size: 29, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:12:35,171 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225457.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:36,591 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225459.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:40,499 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225465.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:44,913 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7479, 2.2924, 4.0086, 1.5688, 2.9574, 2.3713, 1.8556, 2.9702], + device='cuda:1'), covar=tensor([0.1976, 0.3044, 0.0885, 0.5084, 0.1957, 0.3375, 0.2664, 0.2483], + device='cuda:1'), in_proj_covar=tensor([0.0538, 0.0638, 0.0563, 0.0673, 0.0663, 0.0614, 0.0563, 0.0647], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:12:46,185 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225473.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:51,110 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225480.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:53,925 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225484.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:56,557 INFO [train.py:901] (1/4) Epoch 28, batch 7250, loss[loss=0.1716, simple_loss=0.2465, pruned_loss=0.0484, over 7422.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2833, pruned_loss=0.05854, over 1610752.48 frames. ], batch size: 17, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:13:03,596 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225498.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:13:07,602 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225504.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:13:16,360 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.440e+02 2.486e+02 3.022e+02 3.617e+02 8.325e+02, threshold=6.044e+02, percent-clipped=6.0 +2023-02-09 01:13:21,985 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225523.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:13:32,777 INFO [train.py:901] (1/4) Epoch 28, batch 7300, loss[loss=0.1863, simple_loss=0.2719, pruned_loss=0.05032, over 8106.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2826, pruned_loss=0.05836, over 1610361.63 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:13:34,181 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225540.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:13:56,385 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-02-09 01:14:00,391 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225577.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:14:02,407 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225580.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:14:07,681 INFO [train.py:901] (1/4) Epoch 28, batch 7350, loss[loss=0.1762, simple_loss=0.2674, pruned_loss=0.04255, over 8199.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2832, pruned_loss=0.05825, over 1614690.48 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:14:24,562 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-02-09 01:14:27,903 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.502e+02 2.380e+02 2.753e+02 3.463e+02 7.224e+02, threshold=5.506e+02, percent-clipped=3.0 +2023-02-09 01:14:29,522 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225619.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:14:42,948 INFO [train.py:901] (1/4) Epoch 28, batch 7400, loss[loss=0.2312, simple_loss=0.3117, pruned_loss=0.07532, over 8515.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.2841, pruned_loss=0.0588, over 1614644.66 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:14:42,988 WARNING [train.py:1067] (1/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-02-09 01:14:48,673 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8986, 1.3675, 3.4473, 1.5843, 2.4874, 3.7593, 3.8671, 3.2414], + device='cuda:1'), covar=tensor([0.1294, 0.1998, 0.0307, 0.2090, 0.0949, 0.0221, 0.0553, 0.0510], + device='cuda:1'), in_proj_covar=tensor([0.0309, 0.0327, 0.0295, 0.0325, 0.0326, 0.0277, 0.0443, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 01:15:18,721 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.3690, 1.4273, 4.5486, 1.7071, 4.0674, 3.7875, 4.1112, 3.9935], + device='cuda:1'), covar=tensor([0.0638, 0.4786, 0.0529, 0.4489, 0.1040, 0.0994, 0.0637, 0.0674], + device='cuda:1'), in_proj_covar=tensor([0.0683, 0.0664, 0.0735, 0.0660, 0.0747, 0.0636, 0.0645, 0.0717], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:15:19,316 INFO [train.py:901] (1/4) Epoch 28, batch 7450, loss[loss=0.205, simple_loss=0.2802, pruned_loss=0.06495, over 8512.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2836, pruned_loss=0.05822, over 1617414.04 frames. ], batch size: 28, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:15:23,690 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8564, 1.4565, 1.7034, 1.3191, 0.8566, 1.4468, 1.6096, 1.4301], + device='cuda:1'), covar=tensor([0.0610, 0.1292, 0.1649, 0.1528, 0.0639, 0.1546, 0.0762, 0.0714], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0160, 0.0101, 0.0162, 0.0113, 0.0146], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:15:25,007 WARNING [train.py:1067] (1/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-02-09 01:15:40,020 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.685e+02 2.396e+02 3.007e+02 3.866e+02 7.466e+02, threshold=6.014e+02, percent-clipped=6.0 +2023-02-09 01:15:54,482 INFO [train.py:901] (1/4) Epoch 28, batch 7500, loss[loss=0.2037, simple_loss=0.2903, pruned_loss=0.05849, over 8504.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2828, pruned_loss=0.05763, over 1618491.53 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:16:32,496 INFO [train.py:901] (1/4) Epoch 28, batch 7550, loss[loss=0.1883, simple_loss=0.2716, pruned_loss=0.05246, over 7805.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2827, pruned_loss=0.05769, over 1616623.74 frames. ], batch size: 20, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:16:39,079 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6091, 2.5587, 1.8263, 2.3047, 2.2754, 1.5783, 2.2017, 2.3069], + device='cuda:1'), covar=tensor([0.1512, 0.0444, 0.1315, 0.0728, 0.0759, 0.1695, 0.0935, 0.0925], + device='cuda:1'), in_proj_covar=tensor([0.0356, 0.0244, 0.0341, 0.0312, 0.0300, 0.0345, 0.0348, 0.0320], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 01:16:41,874 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225801.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:16:52,725 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.729e+02 2.388e+02 3.127e+02 4.485e+02 1.321e+03, threshold=6.254e+02, percent-clipped=11.0 +2023-02-09 01:16:57,629 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225824.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:05,945 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225836.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:07,143 INFO [train.py:901] (1/4) Epoch 28, batch 7600, loss[loss=0.229, simple_loss=0.3036, pruned_loss=0.07726, over 7202.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2812, pruned_loss=0.0571, over 1613812.74 frames. ], batch size: 72, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:17:23,280 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225861.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:27,273 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225867.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:34,159 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225875.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:37,623 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225880.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:40,887 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225884.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:43,480 INFO [train.py:901] (1/4) Epoch 28, batch 7650, loss[loss=0.2333, simple_loss=0.3158, pruned_loss=0.0754, over 8503.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2825, pruned_loss=0.05771, over 1616027.30 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:17:51,603 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225900.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:01,180 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225913.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:03,341 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225916.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:03,835 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.518e+02 2.353e+02 2.789e+02 3.444e+02 7.654e+02, threshold=5.579e+02, percent-clipped=1.0 +2023-02-09 01:18:06,734 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225921.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:18:18,657 INFO [train.py:901] (1/4) Epoch 28, batch 7700, loss[loss=0.2104, simple_loss=0.2949, pruned_loss=0.06298, over 8502.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2824, pruned_loss=0.05749, over 1617082.98 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:18:19,535 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225939.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:31,341 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-02-09 01:18:44,239 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-02-09 01:18:44,498 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1811, 1.8964, 2.3376, 2.0163, 2.2997, 2.1941, 2.1226, 1.1985], + device='cuda:1'), covar=tensor([0.5704, 0.5020, 0.2238, 0.4272, 0.2647, 0.3639, 0.2069, 0.5580], + device='cuda:1'), in_proj_covar=tensor([0.0964, 0.1031, 0.0836, 0.0999, 0.1025, 0.0936, 0.0772, 0.0855], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 01:18:49,257 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225982.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:53,157 INFO [train.py:901] (1/4) Epoch 28, batch 7750, loss[loss=0.1796, simple_loss=0.252, pruned_loss=0.05362, over 7537.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2817, pruned_loss=0.05728, over 1612558.87 frames. ], batch size: 18, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:19:01,892 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225999.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:19:15,711 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.697e+02 2.439e+02 2.815e+02 3.514e+02 7.333e+02, threshold=5.630e+02, percent-clipped=1.0 +2023-02-09 01:19:29,663 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226036.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:19:30,737 INFO [train.py:901] (1/4) Epoch 28, batch 7800, loss[loss=0.221, simple_loss=0.3007, pruned_loss=0.07066, over 8365.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2826, pruned_loss=0.05777, over 1612618.11 frames. ], batch size: 24, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:19:45,497 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=226059.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:20:05,553 INFO [train.py:901] (1/4) Epoch 28, batch 7850, loss[loss=0.1852, simple_loss=0.2793, pruned_loss=0.04552, over 8557.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2819, pruned_loss=0.0572, over 1612788.47 frames. ], batch size: 49, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:20:25,265 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.578e+02 2.405e+02 3.030e+02 3.828e+02 1.208e+03, threshold=6.060e+02, percent-clipped=4.0 +2023-02-09 01:20:39,665 INFO [train.py:901] (1/4) Epoch 28, batch 7900, loss[loss=0.1912, simple_loss=0.2677, pruned_loss=0.05729, over 7539.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2818, pruned_loss=0.0571, over 1614936.44 frames. ], batch size: 18, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:21:02,886 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226172.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:13,381 INFO [train.py:901] (1/4) Epoch 28, batch 7950, loss[loss=0.2241, simple_loss=0.2979, pruned_loss=0.07514, over 8526.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2832, pruned_loss=0.05843, over 1615826.55 frames. ], batch size: 29, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:21:18,307 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226195.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:19,690 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226197.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:33,033 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.821e+02 2.474e+02 2.920e+02 3.612e+02 7.690e+02, threshold=5.839e+02, percent-clipped=4.0 +2023-02-09 01:21:35,313 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226220.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:36,607 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.6988, 1.5245, 1.8921, 1.4947, 0.9918, 1.5310, 2.1294, 2.1314], + device='cuda:1'), covar=tensor([0.0516, 0.1267, 0.1688, 0.1486, 0.0661, 0.1524, 0.0687, 0.0604], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0161, 0.0102, 0.0163, 0.0113, 0.0146], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:21:37,935 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226224.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:47,691 INFO [train.py:901] (1/4) Epoch 28, batch 8000, loss[loss=0.1858, simple_loss=0.2786, pruned_loss=0.04651, over 8753.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2823, pruned_loss=0.05764, over 1615090.82 frames. ], batch size: 30, lr: 2.66e-03, grad_scale: 16.0 +2023-02-09 01:21:47,901 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226238.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:52,830 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.2483, 1.2622, 3.4030, 1.1220, 2.9754, 2.8547, 3.0878, 2.9984], + device='cuda:1'), covar=tensor([0.0871, 0.4569, 0.0810, 0.4448, 0.1372, 0.1160, 0.0801, 0.0943], + device='cuda:1'), in_proj_covar=tensor([0.0688, 0.0668, 0.0738, 0.0664, 0.0753, 0.0641, 0.0648, 0.0721], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:21:57,007 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=226251.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:59,697 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226255.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:00,947 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226257.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:05,095 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226263.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:17,270 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226280.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:22,557 INFO [train.py:901] (1/4) Epoch 28, batch 8050, loss[loss=0.1933, simple_loss=0.2678, pruned_loss=0.05941, over 7561.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2815, pruned_loss=0.05767, over 1598682.53 frames. ], batch size: 18, lr: 2.66e-03, grad_scale: 16.0 +2023-02-09 01:22:25,397 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226292.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:22:30,115 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.3853, 1.5715, 4.6016, 1.7221, 4.0940, 3.7898, 4.1811, 4.0590], + device='cuda:1'), covar=tensor([0.0597, 0.4339, 0.0510, 0.4347, 0.1080, 0.1027, 0.0522, 0.0679], + device='cuda:1'), in_proj_covar=tensor([0.0685, 0.0665, 0.0735, 0.0661, 0.0749, 0.0638, 0.0644, 0.0718], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:22:37,471 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-02-09 01:22:42,673 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.612e+02 2.434e+02 3.095e+02 3.696e+02 6.520e+02, threshold=6.190e+02, percent-clipped=3.0 +2023-02-09 01:22:42,888 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226317.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:22:57,851 WARNING [train.py:1067] (1/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-09 01:23:01,703 INFO [train.py:901] (1/4) Epoch 29, batch 0, loss[loss=0.1891, simple_loss=0.274, pruned_loss=0.05211, over 7929.00 frames. ], tot_loss[loss=0.1891, simple_loss=0.274, pruned_loss=0.05211, over 7929.00 frames. ], batch size: 20, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:23:01,703 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 01:23:13,264 INFO [train.py:935] (1/4) Epoch 29, validation: loss=0.1705, simple_loss=0.2705, pruned_loss=0.03528, over 944034.00 frames. +2023-02-09 01:23:13,265 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 01:23:26,225 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226339.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:23:29,655 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-02-09 01:23:39,590 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.90 vs. limit=5.0 +2023-02-09 01:23:49,914 INFO [train.py:901] (1/4) Epoch 29, batch 50, loss[loss=0.2237, simple_loss=0.3086, pruned_loss=0.06942, over 8769.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2855, pruned_loss=0.06025, over 368168.85 frames. ], batch size: 40, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:23:50,830 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226372.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:23:53,661 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7005, 1.9195, 1.9411, 1.4428, 2.0929, 1.4167, 0.5997, 1.9377], + device='cuda:1'), covar=tensor([0.0660, 0.0430, 0.0361, 0.0603, 0.0452, 0.1049, 0.1091, 0.0319], + device='cuda:1'), in_proj_covar=tensor([0.0478, 0.0414, 0.0368, 0.0462, 0.0398, 0.0555, 0.0406, 0.0443], + device='cuda:1'), out_proj_covar=tensor([1.2657e-04, 1.0724e-04, 9.5925e-05, 1.2073e-04, 1.0402e-04, 1.5462e-04, + 1.0826e-04, 1.1606e-04], device='cuda:1') +2023-02-09 01:24:06,019 WARNING [train.py:1067] (1/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-02-09 01:24:12,517 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226403.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:24:22,934 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.399e+02 2.293e+02 2.936e+02 3.721e+02 6.222e+02, threshold=5.872e+02, percent-clipped=1.0 +2023-02-09 01:24:25,777 INFO [train.py:901] (1/4) Epoch 29, batch 100, loss[loss=0.1604, simple_loss=0.2575, pruned_loss=0.03164, over 8252.00 frames. ], tot_loss[loss=0.2037, simple_loss=0.2874, pruned_loss=0.05998, over 648804.17 frames. ], batch size: 24, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:24:30,614 WARNING [train.py:1067] (1/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-02-09 01:24:33,769 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1517, 1.9584, 2.4465, 2.0914, 2.3912, 2.2323, 2.0838, 1.3477], + device='cuda:1'), covar=tensor([0.5865, 0.5081, 0.2267, 0.4116, 0.2792, 0.3427, 0.2031, 0.5554], + device='cuda:1'), in_proj_covar=tensor([0.0969, 0.1034, 0.0840, 0.1003, 0.1030, 0.0940, 0.0776, 0.0858], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 01:25:02,686 INFO [train.py:901] (1/4) Epoch 29, batch 150, loss[loss=0.1671, simple_loss=0.2458, pruned_loss=0.04416, over 7800.00 frames. ], tot_loss[loss=0.2047, simple_loss=0.2873, pruned_loss=0.06105, over 860979.21 frames. ], batch size: 19, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:25:34,579 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.592e+02 2.438e+02 2.916e+02 4.111e+02 7.524e+02, threshold=5.832e+02, percent-clipped=2.0 +2023-02-09 01:25:35,513 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226518.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:25:37,496 INFO [train.py:901] (1/4) Epoch 29, batch 200, loss[loss=0.1507, simple_loss=0.2344, pruned_loss=0.03355, over 7687.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.286, pruned_loss=0.06008, over 1027035.00 frames. ], batch size: 18, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:26:12,529 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.1822, 1.3298, 1.6916, 1.3065, 0.7367, 1.4611, 1.2196, 1.0664], + device='cuda:1'), covar=tensor([0.0658, 0.1310, 0.1672, 0.1477, 0.0580, 0.1483, 0.0696, 0.0745], + device='cuda:1'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0161, 0.0102, 0.0163, 0.0113, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:26:12,615 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0023, 1.7448, 2.0307, 1.8333, 1.9825, 2.0665, 1.9610, 0.9098], + device='cuda:1'), covar=tensor([0.5906, 0.5143, 0.2409, 0.4039, 0.2604, 0.3538, 0.2069, 0.5407], + device='cuda:1'), in_proj_covar=tensor([0.0965, 0.1030, 0.0837, 0.0999, 0.1025, 0.0935, 0.0773, 0.0852], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 01:26:13,673 INFO [train.py:901] (1/4) Epoch 29, batch 250, loss[loss=0.1976, simple_loss=0.2815, pruned_loss=0.05689, over 8321.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2848, pruned_loss=0.05925, over 1159601.88 frames. ], batch size: 25, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:26:26,058 WARNING [train.py:1067] (1/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-02-09 01:26:31,141 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226595.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:26:31,292 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226595.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:26:33,830 WARNING [train.py:1067] (1/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-02-09 01:26:46,400 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.753e+02 2.514e+02 3.003e+02 3.645e+02 8.891e+02, threshold=6.006e+02, percent-clipped=9.0 +2023-02-09 01:26:48,779 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226620.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:26:49,245 INFO [train.py:901] (1/4) Epoch 29, batch 300, loss[loss=0.2207, simple_loss=0.2999, pruned_loss=0.07076, over 8345.00 frames. ], tot_loss[loss=0.2013, simple_loss=0.2846, pruned_loss=0.059, over 1267199.19 frames. ], batch size: 26, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:26:54,267 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226628.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:27:12,774 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226653.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:27:25,791 INFO [train.py:901] (1/4) Epoch 29, batch 350, loss[loss=0.1972, simple_loss=0.2897, pruned_loss=0.05234, over 8489.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2837, pruned_loss=0.05892, over 1342437.11 frames. ], batch size: 29, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:27:45,003 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2152, 2.4861, 2.7637, 1.6010, 3.1946, 1.7945, 1.5835, 2.1695], + device='cuda:1'), covar=tensor([0.0991, 0.0515, 0.0376, 0.0994, 0.0532, 0.1077, 0.1133, 0.0772], + device='cuda:1'), in_proj_covar=tensor([0.0483, 0.0418, 0.0372, 0.0466, 0.0401, 0.0559, 0.0408, 0.0447], + device='cuda:1'), out_proj_covar=tensor([1.2785e-04, 1.0831e-04, 9.7018e-05, 1.2160e-04, 1.0475e-04, 1.5584e-04, + 1.0894e-04, 1.1705e-04], device='cuda:1') +2023-02-09 01:27:46,743 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4106, 1.6284, 2.0587, 1.3741, 1.4570, 1.6930, 1.4946, 1.4616], + device='cuda:1'), covar=tensor([0.2030, 0.2704, 0.1106, 0.4664, 0.2238, 0.3566, 0.2616, 0.2289], + device='cuda:1'), in_proj_covar=tensor([0.0540, 0.0637, 0.0562, 0.0671, 0.0664, 0.0614, 0.0564, 0.0646], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:27:54,190 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226710.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:27:58,932 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.666e+02 2.400e+02 2.862e+02 3.557e+02 6.632e+02, threshold=5.725e+02, percent-clipped=2.0 +2023-02-09 01:28:01,712 INFO [train.py:901] (1/4) Epoch 29, batch 400, loss[loss=0.1865, simple_loss=0.2779, pruned_loss=0.04753, over 8094.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2839, pruned_loss=0.05831, over 1406346.10 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:28:37,571 INFO [train.py:901] (1/4) Epoch 29, batch 450, loss[loss=0.1422, simple_loss=0.2252, pruned_loss=0.02963, over 7803.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2832, pruned_loss=0.05756, over 1452706.22 frames. ], batch size: 19, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:28:38,565 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=1.99 vs. limit=5.0 +2023-02-09 01:28:39,871 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226774.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:28:53,563 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5361, 1.4623, 1.8303, 1.2147, 1.1548, 1.8367, 0.3182, 1.2536], + device='cuda:1'), covar=tensor([0.1430, 0.1111, 0.0366, 0.0739, 0.2285, 0.0397, 0.1696, 0.1023], + device='cuda:1'), in_proj_covar=tensor([0.0202, 0.0208, 0.0138, 0.0224, 0.0280, 0.0148, 0.0174, 0.0200], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 01:28:56,601 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-02-09 01:28:58,207 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226799.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:29:11,132 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.602e+02 2.449e+02 2.964e+02 3.856e+02 9.700e+02, threshold=5.929e+02, percent-clipped=9.0 +2023-02-09 01:29:13,796 INFO [train.py:901] (1/4) Epoch 29, batch 500, loss[loss=0.2672, simple_loss=0.3269, pruned_loss=0.1037, over 6707.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2827, pruned_loss=0.05755, over 1488248.61 frames. ], batch size: 72, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:29:48,217 INFO [train.py:901] (1/4) Epoch 29, batch 550, loss[loss=0.1997, simple_loss=0.2839, pruned_loss=0.05779, over 8286.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2824, pruned_loss=0.05719, over 1518498.43 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:29:51,497 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-02-09 01:30:21,906 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.374e+02 2.486e+02 3.156e+02 4.092e+02 1.034e+03, threshold=6.313e+02, percent-clipped=6.0 +2023-02-09 01:30:24,623 INFO [train.py:901] (1/4) Epoch 29, batch 600, loss[loss=0.1872, simple_loss=0.2647, pruned_loss=0.05489, over 7966.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2831, pruned_loss=0.0575, over 1544731.24 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:30:28,540 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.68 vs. limit=5.0 +2023-02-09 01:30:35,116 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4575, 2.0329, 4.6417, 2.1331, 2.6919, 5.1658, 5.3467, 4.5549], + device='cuda:1'), covar=tensor([0.1208, 0.1662, 0.0194, 0.1901, 0.1037, 0.0175, 0.0262, 0.0515], + device='cuda:1'), in_proj_covar=tensor([0.0310, 0.0327, 0.0295, 0.0325, 0.0327, 0.0278, 0.0444, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 01:30:38,671 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.6248, 2.3991, 3.0892, 2.5560, 3.1225, 2.6279, 2.5479, 1.9046], + device='cuda:1'), covar=tensor([0.5799, 0.5635, 0.2290, 0.4229, 0.2789, 0.3108, 0.1782, 0.5928], + device='cuda:1'), in_proj_covar=tensor([0.0965, 0.1033, 0.0838, 0.0999, 0.1026, 0.0936, 0.0772, 0.0852], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 01:30:43,053 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-02-09 01:30:56,503 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226966.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:30:59,753 INFO [train.py:901] (1/4) Epoch 29, batch 650, loss[loss=0.199, simple_loss=0.2932, pruned_loss=0.05238, over 8360.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2825, pruned_loss=0.05773, over 1554561.09 frames. ], batch size: 26, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:31:13,843 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226991.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:31:25,429 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227007.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:31:32,762 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.693e+02 2.425e+02 2.942e+02 3.859e+02 6.314e+02, threshold=5.885e+02, percent-clipped=1.0 +2023-02-09 01:31:36,261 INFO [train.py:901] (1/4) Epoch 29, batch 700, loss[loss=0.2022, simple_loss=0.2996, pruned_loss=0.05238, over 8513.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2829, pruned_loss=0.05778, over 1563043.17 frames. ], batch size: 39, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:31:56,845 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9901, 2.2028, 1.8207, 2.8653, 1.4800, 1.6873, 2.2431, 2.2785], + device='cuda:1'), covar=tensor([0.0927, 0.0794, 0.1090, 0.0442, 0.1040, 0.1451, 0.0774, 0.0793], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0196, 0.0245, 0.0214, 0.0204, 0.0249, 0.0251, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 01:32:12,699 INFO [train.py:901] (1/4) Epoch 29, batch 750, loss[loss=0.2129, simple_loss=0.2977, pruned_loss=0.06407, over 8348.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2831, pruned_loss=0.05779, over 1576683.50 frames. ], batch size: 26, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:32:31,396 WARNING [train.py:1067] (1/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-02-09 01:32:40,246 WARNING [train.py:1067] (1/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-02-09 01:32:44,348 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.694e+02 2.547e+02 3.055e+02 4.023e+02 1.198e+03, threshold=6.109e+02, percent-clipped=3.0 +2023-02-09 01:32:47,760 INFO [train.py:901] (1/4) Epoch 29, batch 800, loss[loss=0.1892, simple_loss=0.2596, pruned_loss=0.05939, over 7707.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2827, pruned_loss=0.05771, over 1584632.88 frames. ], batch size: 18, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:33:24,580 INFO [train.py:901] (1/4) Epoch 29, batch 850, loss[loss=0.1964, simple_loss=0.2782, pruned_loss=0.05731, over 8245.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.281, pruned_loss=0.05656, over 1585892.69 frames. ], batch size: 22, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:33:57,024 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.463e+02 2.886e+02 3.949e+02 8.845e+02, threshold=5.773e+02, percent-clipped=3.0 +2023-02-09 01:33:59,146 INFO [train.py:901] (1/4) Epoch 29, batch 900, loss[loss=0.2122, simple_loss=0.2836, pruned_loss=0.07035, over 8070.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2804, pruned_loss=0.05628, over 1593686.63 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:34:35,613 INFO [train.py:901] (1/4) Epoch 29, batch 950, loss[loss=0.2221, simple_loss=0.3004, pruned_loss=0.07194, over 8300.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2816, pruned_loss=0.05677, over 1601952.00 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:35:04,861 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-02-09 01:35:09,005 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.642e+02 2.581e+02 3.033e+02 3.905e+02 1.035e+03, threshold=6.066e+02, percent-clipped=4.0 +2023-02-09 01:35:11,178 INFO [train.py:901] (1/4) Epoch 29, batch 1000, loss[loss=0.1891, simple_loss=0.2601, pruned_loss=0.05903, over 7533.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.282, pruned_loss=0.05671, over 1610323.46 frames. ], batch size: 18, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:35:32,312 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=227351.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:35:39,825 WARNING [train.py:1067] (1/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-02-09 01:35:47,252 INFO [train.py:901] (1/4) Epoch 29, batch 1050, loss[loss=0.1921, simple_loss=0.2673, pruned_loss=0.05844, over 5966.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2815, pruned_loss=0.05619, over 1609886.90 frames. ], batch size: 13, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:35:52,716 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-02-09 01:36:22,051 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.565e+02 2.373e+02 3.020e+02 3.651e+02 1.051e+03, threshold=6.040e+02, percent-clipped=1.0 +2023-02-09 01:36:23,015 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2415, 1.0778, 1.3289, 1.0621, 1.0117, 1.3375, 0.1632, 1.0650], + device='cuda:1'), covar=tensor([0.1385, 0.1299, 0.0489, 0.0646, 0.2302, 0.0542, 0.1733, 0.1071], + device='cuda:1'), in_proj_covar=tensor([0.0202, 0.0207, 0.0137, 0.0224, 0.0278, 0.0147, 0.0173, 0.0199], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 01:36:24,319 INFO [train.py:901] (1/4) Epoch 29, batch 1100, loss[loss=0.143, simple_loss=0.2283, pruned_loss=0.02888, over 7707.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2819, pruned_loss=0.05634, over 1616144.86 frames. ], batch size: 18, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:36:30,384 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227429.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:36:56,358 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=227466.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:36:59,076 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8390, 1.4732, 1.6590, 1.4191, 1.0318, 1.5080, 1.7080, 1.5099], + device='cuda:1'), covar=tensor([0.0585, 0.1354, 0.1823, 0.1570, 0.0631, 0.1581, 0.0744, 0.0713], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0154, 0.0191, 0.0163, 0.0102, 0.0164, 0.0114, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:36:59,547 INFO [train.py:901] (1/4) Epoch 29, batch 1150, loss[loss=0.1808, simple_loss=0.2756, pruned_loss=0.04296, over 8293.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2809, pruned_loss=0.05619, over 1618672.02 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:37:06,458 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-02-09 01:37:34,273 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.484e+02 2.320e+02 2.698e+02 3.625e+02 7.425e+02, threshold=5.396e+02, percent-clipped=3.0 +2023-02-09 01:37:36,428 INFO [train.py:901] (1/4) Epoch 29, batch 1200, loss[loss=0.1636, simple_loss=0.2498, pruned_loss=0.03873, over 7816.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2803, pruned_loss=0.05556, over 1617940.64 frames. ], batch size: 20, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:37:53,824 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1561, 2.2698, 1.9404, 2.8523, 1.2663, 1.7711, 1.9615, 2.3055], + device='cuda:1'), covar=tensor([0.0695, 0.0775, 0.0792, 0.0330, 0.1197, 0.1250, 0.0905, 0.0804], + device='cuda:1'), in_proj_covar=tensor([0.0230, 0.0195, 0.0244, 0.0213, 0.0203, 0.0247, 0.0250, 0.0205], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 01:38:09,053 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8719, 1.5471, 1.7819, 1.4706, 0.9704, 1.5400, 1.6811, 1.4977], + device='cuda:1'), covar=tensor([0.0587, 0.1235, 0.1634, 0.1411, 0.0610, 0.1425, 0.0707, 0.0676], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0154, 0.0190, 0.0162, 0.0102, 0.0163, 0.0113, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:38:11,541 INFO [train.py:901] (1/4) Epoch 29, batch 1250, loss[loss=0.1869, simple_loss=0.2741, pruned_loss=0.04981, over 8191.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2814, pruned_loss=0.05648, over 1618493.09 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:38:13,782 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4579, 3.8687, 2.6931, 3.3560, 3.2007, 2.3494, 3.1956, 3.4442], + device='cuda:1'), covar=tensor([0.1594, 0.0377, 0.1044, 0.0663, 0.0667, 0.1437, 0.0976, 0.0974], + device='cuda:1'), in_proj_covar=tensor([0.0356, 0.0246, 0.0344, 0.0314, 0.0303, 0.0349, 0.0351, 0.0323], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 01:38:19,535 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7373, 2.1221, 3.2784, 1.6166, 2.5788, 2.1906, 1.8539, 2.5458], + device='cuda:1'), covar=tensor([0.1895, 0.2801, 0.0914, 0.4806, 0.1864, 0.3390, 0.2440, 0.2367], + device='cuda:1'), in_proj_covar=tensor([0.0539, 0.0639, 0.0564, 0.0672, 0.0667, 0.0617, 0.0565, 0.0648], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:38:46,455 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.451e+02 2.398e+02 2.828e+02 3.393e+02 7.704e+02, threshold=5.657e+02, percent-clipped=4.0 +2023-02-09 01:38:48,671 INFO [train.py:901] (1/4) Epoch 29, batch 1300, loss[loss=0.2198, simple_loss=0.3081, pruned_loss=0.06578, over 8449.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2812, pruned_loss=0.05602, over 1620466.34 frames. ], batch size: 27, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:39:24,554 INFO [train.py:901] (1/4) Epoch 29, batch 1350, loss[loss=0.1999, simple_loss=0.2919, pruned_loss=0.05393, over 8674.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2814, pruned_loss=0.05627, over 1621382.94 frames. ], batch size: 49, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:39:52,814 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227711.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:39:58,210 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.624e+02 2.335e+02 2.785e+02 3.650e+02 1.055e+03, threshold=5.570e+02, percent-clipped=4.0 +2023-02-09 01:40:00,377 INFO [train.py:901] (1/4) Epoch 29, batch 1400, loss[loss=0.2189, simple_loss=0.3099, pruned_loss=0.06398, over 8290.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2808, pruned_loss=0.05572, over 1622592.95 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:40:01,311 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=227722.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:40:20,607 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=227747.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:40:38,113 INFO [train.py:901] (1/4) Epoch 29, batch 1450, loss[loss=0.1838, simple_loss=0.2776, pruned_loss=0.04495, over 8355.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2807, pruned_loss=0.05538, over 1621407.37 frames. ], batch size: 26, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:40:39,477 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=227773.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:40:47,997 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-02-09 01:41:11,911 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.751e+02 2.369e+02 2.795e+02 3.434e+02 1.018e+03, threshold=5.589e+02, percent-clipped=3.0 +2023-02-09 01:41:14,105 INFO [train.py:901] (1/4) Epoch 29, batch 1500, loss[loss=0.1726, simple_loss=0.2646, pruned_loss=0.0403, over 8463.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2799, pruned_loss=0.05533, over 1620636.75 frames. ], batch size: 25, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:41:31,878 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227845.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:41:40,138 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.6252, 4.6222, 4.0775, 1.8584, 4.0594, 4.2421, 4.1747, 4.0461], + device='cuda:1'), covar=tensor([0.0632, 0.0424, 0.0978, 0.4442, 0.0871, 0.0869, 0.1065, 0.0681], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0458, 0.0456, 0.0563, 0.0446, 0.0470, 0.0442, 0.0414], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:41:51,485 INFO [train.py:901] (1/4) Epoch 29, batch 1550, loss[loss=0.1861, simple_loss=0.2786, pruned_loss=0.04682, over 8075.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2813, pruned_loss=0.05639, over 1622372.21 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:42:04,411 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=227888.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:42:25,548 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.467e+02 2.452e+02 3.275e+02 4.864e+02 1.208e+03, threshold=6.551e+02, percent-clipped=17.0 +2023-02-09 01:42:27,699 INFO [train.py:901] (1/4) Epoch 29, batch 1600, loss[loss=0.1814, simple_loss=0.2853, pruned_loss=0.03875, over 8339.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2813, pruned_loss=0.05615, over 1622706.08 frames. ], batch size: 26, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:42:27,875 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4477, 1.7912, 1.4219, 3.0424, 1.3541, 1.2968, 2.1110, 2.0181], + device='cuda:1'), covar=tensor([0.1597, 0.1418, 0.1960, 0.0340, 0.1453, 0.2243, 0.1067, 0.1122], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0195, 0.0245, 0.0213, 0.0204, 0.0247, 0.0251, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 01:43:04,958 INFO [train.py:901] (1/4) Epoch 29, batch 1650, loss[loss=0.1979, simple_loss=0.2912, pruned_loss=0.05237, over 8582.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2808, pruned_loss=0.056, over 1619343.12 frames. ], batch size: 31, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:43:39,886 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.526e+02 2.401e+02 2.687e+02 3.396e+02 6.045e+02, threshold=5.374e+02, percent-clipped=0.0 +2023-02-09 01:43:42,081 INFO [train.py:901] (1/4) Epoch 29, batch 1700, loss[loss=0.2185, simple_loss=0.303, pruned_loss=0.06703, over 8195.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2805, pruned_loss=0.05594, over 1622456.93 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:43:42,948 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.6441, 4.6718, 4.1673, 2.4000, 4.1236, 4.2451, 4.2492, 4.0609], + device='cuda:1'), covar=tensor([0.0682, 0.0485, 0.0975, 0.4110, 0.0864, 0.1078, 0.1058, 0.0832], + device='cuda:1'), in_proj_covar=tensor([0.0548, 0.0461, 0.0457, 0.0567, 0.0447, 0.0472, 0.0445, 0.0415], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:43:44,491 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1843, 1.7744, 4.1368, 1.8607, 2.4363, 4.6413, 4.7830, 4.0807], + device='cuda:1'), covar=tensor([0.1231, 0.1849, 0.0308, 0.2060, 0.1262, 0.0209, 0.0476, 0.0526], + device='cuda:1'), in_proj_covar=tensor([0.0312, 0.0329, 0.0298, 0.0327, 0.0329, 0.0280, 0.0450, 0.0312], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 01:43:45,195 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228025.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:44:06,282 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228055.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:44:18,095 INFO [train.py:901] (1/4) Epoch 29, batch 1750, loss[loss=0.1941, simple_loss=0.2678, pruned_loss=0.06017, over 7780.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2802, pruned_loss=0.05592, over 1623144.98 frames. ], batch size: 19, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:44:28,126 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1567, 2.0646, 2.5913, 2.2599, 2.6118, 2.2685, 2.1926, 1.5961], + device='cuda:1'), covar=tensor([0.6069, 0.5174, 0.2199, 0.3826, 0.2579, 0.3276, 0.2047, 0.5303], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1038, 0.0842, 0.1005, 0.1030, 0.0942, 0.0776, 0.0857], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 01:44:38,632 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228098.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:44:38,667 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0040, 1.6005, 1.8779, 1.4428, 1.1310, 1.5647, 1.8545, 1.6692], + device='cuda:1'), covar=tensor([0.0555, 0.1235, 0.1579, 0.1471, 0.0569, 0.1416, 0.0660, 0.0648], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0154, 0.0190, 0.0162, 0.0102, 0.0164, 0.0114, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:44:52,743 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.427e+02 2.425e+02 2.925e+02 3.463e+02 6.679e+02, threshold=5.849e+02, percent-clipped=2.0 +2023-02-09 01:44:55,484 INFO [train.py:901] (1/4) Epoch 29, batch 1800, loss[loss=0.2151, simple_loss=0.3081, pruned_loss=0.06109, over 8098.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2815, pruned_loss=0.05663, over 1618086.45 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:45:11,244 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228144.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:45:12,123 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.76 vs. limit=2.0 +2023-02-09 01:45:28,614 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228169.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:45:29,268 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228170.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:45:29,726 INFO [train.py:901] (1/4) Epoch 29, batch 1850, loss[loss=0.197, simple_loss=0.2786, pruned_loss=0.05776, over 8593.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.282, pruned_loss=0.05707, over 1616382.46 frames. ], batch size: 31, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:45:42,839 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228189.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:46:03,761 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.759e+02 2.509e+02 2.932e+02 3.424e+02 5.958e+02, threshold=5.864e+02, percent-clipped=1.0 +2023-02-09 01:46:05,868 INFO [train.py:901] (1/4) Epoch 29, batch 1900, loss[loss=0.1774, simple_loss=0.2642, pruned_loss=0.04527, over 7970.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2813, pruned_loss=0.05635, over 1618586.00 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:46:29,135 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.36 vs. limit=5.0 +2023-02-09 01:46:32,890 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.36 vs. limit=2.0 +2023-02-09 01:46:38,539 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-02-09 01:46:41,415 INFO [train.py:901] (1/4) Epoch 29, batch 1950, loss[loss=0.2199, simple_loss=0.3034, pruned_loss=0.0682, over 8690.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.282, pruned_loss=0.05646, over 1622366.19 frames. ], batch size: 34, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:46:50,971 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-02-09 01:47:05,322 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228304.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:10,131 WARNING [train.py:1067] (1/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-02-09 01:47:16,244 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.886e+02 2.422e+02 3.056e+02 3.762e+02 8.552e+02, threshold=6.111e+02, percent-clipped=4.0 +2023-02-09 01:47:18,252 INFO [train.py:901] (1/4) Epoch 29, batch 2000, loss[loss=0.1751, simple_loss=0.2473, pruned_loss=0.05149, over 7241.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2811, pruned_loss=0.05607, over 1624090.06 frames. ], batch size: 16, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:47:36,281 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228347.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:39,103 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228351.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:44,563 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.1807, 1.3007, 1.3056, 0.9218, 1.3238, 1.1009, 0.3511, 1.2935], + device='cuda:1'), covar=tensor([0.0438, 0.0329, 0.0256, 0.0441, 0.0368, 0.0634, 0.0775, 0.0257], + device='cuda:1'), in_proj_covar=tensor([0.0476, 0.0414, 0.0369, 0.0460, 0.0397, 0.0552, 0.0403, 0.0443], + device='cuda:1'), out_proj_covar=tensor([1.2611e-04, 1.0730e-04, 9.6024e-05, 1.2017e-04, 1.0384e-04, 1.5379e-04, + 1.0767e-04, 1.1589e-04], device='cuda:1') +2023-02-09 01:47:52,202 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228369.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:53,536 INFO [train.py:901] (1/4) Epoch 29, batch 2050, loss[loss=0.2039, simple_loss=0.2866, pruned_loss=0.06062, over 8334.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2808, pruned_loss=0.05612, over 1623533.63 frames. ], batch size: 26, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:48:25,953 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.725e+02 2.333e+02 2.866e+02 3.600e+02 5.490e+02, threshold=5.733e+02, percent-clipped=0.0 +2023-02-09 01:48:28,117 INFO [train.py:901] (1/4) Epoch 29, batch 2100, loss[loss=0.1676, simple_loss=0.2666, pruned_loss=0.03428, over 8200.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2802, pruned_loss=0.05548, over 1621156.71 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:48:32,344 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228426.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:48:43,920 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228442.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:48:48,016 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228447.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:48:50,743 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228451.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:48:59,117 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6430, 2.0662, 3.2715, 1.5759, 2.6430, 2.0871, 1.7608, 2.6882], + device='cuda:1'), covar=tensor([0.2050, 0.2835, 0.0975, 0.4960, 0.1791, 0.3432, 0.2590, 0.2260], + device='cuda:1'), in_proj_covar=tensor([0.0543, 0.0642, 0.0568, 0.0675, 0.0667, 0.0617, 0.0566, 0.0650], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:49:04,298 INFO [train.py:901] (1/4) Epoch 29, batch 2150, loss[loss=0.2003, simple_loss=0.272, pruned_loss=0.06432, over 7648.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2804, pruned_loss=0.05578, over 1620356.52 frames. ], batch size: 19, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:49:05,162 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.8255, 1.3411, 4.0061, 1.4287, 3.5419, 3.3687, 3.6441, 3.5058], + device='cuda:1'), covar=tensor([0.0712, 0.4555, 0.0679, 0.4475, 0.1265, 0.1046, 0.0656, 0.0833], + device='cuda:1'), in_proj_covar=tensor([0.0686, 0.0669, 0.0744, 0.0664, 0.0750, 0.0639, 0.0647, 0.0721], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:49:14,145 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228484.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:49:37,705 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.627e+02 2.659e+02 3.270e+02 3.951e+02 1.171e+03, threshold=6.540e+02, percent-clipped=10.0 +2023-02-09 01:49:39,911 INFO [train.py:901] (1/4) Epoch 29, batch 2200, loss[loss=0.216, simple_loss=0.3037, pruned_loss=0.0641, over 8634.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2801, pruned_loss=0.05574, over 1616397.28 frames. ], batch size: 49, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:49:47,016 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.2464, 1.3054, 3.3856, 1.1213, 3.0016, 2.8459, 3.0868, 3.0001], + device='cuda:1'), covar=tensor([0.0875, 0.4153, 0.0874, 0.4338, 0.1319, 0.1085, 0.0795, 0.0935], + device='cuda:1'), in_proj_covar=tensor([0.0684, 0.0666, 0.0742, 0.0662, 0.0747, 0.0637, 0.0644, 0.0719], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:50:06,272 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228557.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:50:08,439 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228560.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:50:16,499 INFO [train.py:901] (1/4) Epoch 29, batch 2250, loss[loss=0.2292, simple_loss=0.3121, pruned_loss=0.07315, over 8472.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2817, pruned_loss=0.0567, over 1614749.16 frames. ], batch size: 29, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:50:26,600 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228585.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:50:44,232 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4806, 1.3194, 2.3294, 1.2965, 2.3177, 2.4928, 2.6428, 2.0486], + device='cuda:1'), covar=tensor([0.1175, 0.1568, 0.0524, 0.2119, 0.0783, 0.0475, 0.0814, 0.0880], + device='cuda:1'), in_proj_covar=tensor([0.0311, 0.0327, 0.0297, 0.0325, 0.0329, 0.0279, 0.0447, 0.0312], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 01:50:50,350 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.529e+02 2.350e+02 2.877e+02 3.583e+02 6.549e+02, threshold=5.755e+02, percent-clipped=1.0 +2023-02-09 01:50:52,578 INFO [train.py:901] (1/4) Epoch 29, batch 2300, loss[loss=0.1884, simple_loss=0.267, pruned_loss=0.05489, over 7782.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2817, pruned_loss=0.05706, over 1608149.29 frames. ], batch size: 19, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:50:53,470 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6518, 1.5286, 2.1233, 1.3570, 1.2281, 2.0909, 0.3599, 1.2727], + device='cuda:1'), covar=tensor([0.1400, 0.1310, 0.0354, 0.1037, 0.2474, 0.0481, 0.1906, 0.1289], + device='cuda:1'), in_proj_covar=tensor([0.0202, 0.0206, 0.0138, 0.0223, 0.0278, 0.0149, 0.0173, 0.0199], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 01:51:06,297 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.73 vs. limit=2.0 +2023-02-09 01:51:22,792 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.1689, 1.5272, 4.3159, 1.5786, 3.8574, 3.5740, 3.8784, 3.7500], + device='cuda:1'), covar=tensor([0.0650, 0.4711, 0.0610, 0.4373, 0.1162, 0.1081, 0.0623, 0.0760], + device='cuda:1'), in_proj_covar=tensor([0.0686, 0.0669, 0.0746, 0.0666, 0.0751, 0.0640, 0.0647, 0.0722], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:51:28,785 INFO [train.py:901] (1/4) Epoch 29, batch 2350, loss[loss=0.1968, simple_loss=0.2855, pruned_loss=0.0541, over 8326.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2827, pruned_loss=0.05697, over 1616868.90 frames. ], batch size: 26, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:51:43,129 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228691.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:51:45,887 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228695.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:52:02,338 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.786e+02 2.434e+02 3.194e+02 3.873e+02 7.294e+02, threshold=6.388e+02, percent-clipped=4.0 +2023-02-09 01:52:04,285 INFO [train.py:901] (1/4) Epoch 29, batch 2400, loss[loss=0.204, simple_loss=0.2732, pruned_loss=0.06741, over 7976.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2834, pruned_loss=0.05763, over 1619003.17 frames. ], batch size: 21, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:52:17,436 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228740.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:52:34,954 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228765.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:52:39,614 INFO [train.py:901] (1/4) Epoch 29, batch 2450, loss[loss=0.1791, simple_loss=0.2713, pruned_loss=0.04343, over 8288.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.282, pruned_loss=0.05696, over 1619033.67 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:52:55,038 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228791.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:53:05,513 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228806.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:08,257 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228810.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:10,298 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228813.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:13,501 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.470e+02 2.278e+02 2.630e+02 3.527e+02 5.802e+02, threshold=5.259e+02, percent-clipped=0.0 +2023-02-09 01:53:16,198 INFO [train.py:901] (1/4) Epoch 29, batch 2500, loss[loss=0.1809, simple_loss=0.2701, pruned_loss=0.04584, over 8227.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2822, pruned_loss=0.05729, over 1617035.09 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:53:28,187 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228838.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:49,886 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-02-09 01:53:51,011 INFO [train.py:901] (1/4) Epoch 29, batch 2550, loss[loss=0.1984, simple_loss=0.2622, pruned_loss=0.06726, over 7798.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2813, pruned_loss=0.05679, over 1617083.88 frames. ], batch size: 19, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:54:17,016 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228906.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:54:25,705 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.542e+02 2.320e+02 2.761e+02 3.420e+02 6.403e+02, threshold=5.523e+02, percent-clipped=2.0 +2023-02-09 01:54:27,391 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0336, 2.3286, 3.1609, 1.9399, 2.7518, 2.3801, 2.1319, 2.7492], + device='cuda:1'), covar=tensor([0.1464, 0.2053, 0.0685, 0.3384, 0.1332, 0.2453, 0.1864, 0.1736], + device='cuda:1'), in_proj_covar=tensor([0.0546, 0.0644, 0.0569, 0.0679, 0.0670, 0.0619, 0.0568, 0.0652], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:54:27,842 INFO [train.py:901] (1/4) Epoch 29, batch 2600, loss[loss=0.2189, simple_loss=0.2996, pruned_loss=0.06912, over 7011.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2813, pruned_loss=0.05676, over 1618166.41 frames. ], batch size: 71, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:54:51,301 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228953.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:55:03,876 INFO [train.py:901] (1/4) Epoch 29, batch 2650, loss[loss=0.1793, simple_loss=0.2626, pruned_loss=0.04797, over 8092.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2801, pruned_loss=0.05635, over 1614693.86 frames. ], batch size: 21, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:55:27,209 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.2824, 1.3785, 3.3843, 1.1240, 3.0960, 2.8629, 3.1210, 3.0545], + device='cuda:1'), covar=tensor([0.0771, 0.3832, 0.0793, 0.4090, 0.1120, 0.1015, 0.0634, 0.0815], + device='cuda:1'), in_proj_covar=tensor([0.0681, 0.0663, 0.0741, 0.0662, 0.0743, 0.0635, 0.0640, 0.0715], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 01:55:37,362 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.671e+02 2.477e+02 2.922e+02 3.574e+02 8.790e+02, threshold=5.845e+02, percent-clipped=2.0 +2023-02-09 01:55:39,868 INFO [train.py:901] (1/4) Epoch 29, batch 2700, loss[loss=0.2472, simple_loss=0.3166, pruned_loss=0.08895, over 8579.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2803, pruned_loss=0.05632, over 1612815.11 frames. ], batch size: 31, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:56:11,095 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229062.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:14,605 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229066.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:17,767 INFO [train.py:901] (1/4) Epoch 29, batch 2750, loss[loss=0.1735, simple_loss=0.2622, pruned_loss=0.04245, over 8024.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2804, pruned_loss=0.05627, over 1616018.07 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:56:29,425 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229087.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:32,342 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229091.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:51,303 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.676e+02 2.406e+02 2.857e+02 3.824e+02 7.570e+02, threshold=5.715e+02, percent-clipped=1.0 +2023-02-09 01:56:53,439 INFO [train.py:901] (1/4) Epoch 29, batch 2800, loss[loss=0.2118, simple_loss=0.295, pruned_loss=0.06423, over 8294.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2793, pruned_loss=0.05584, over 1610315.29 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:56:54,323 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2034, 1.1416, 1.5429, 1.1130, 0.7870, 1.3254, 1.2780, 1.0628], + device='cuda:1'), covar=tensor([0.0681, 0.1747, 0.2212, 0.1884, 0.0651, 0.1971, 0.0784, 0.0796], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0154, 0.0190, 0.0162, 0.0102, 0.0163, 0.0114, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:57:24,913 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229162.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:57:31,016 INFO [train.py:901] (1/4) Epoch 29, batch 2850, loss[loss=0.2299, simple_loss=0.2972, pruned_loss=0.08124, over 8600.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2805, pruned_loss=0.05652, over 1611720.38 frames. ], batch size: 39, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:57:43,223 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229187.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:58:03,986 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1597, 1.6998, 3.1034, 1.6206, 2.3367, 3.4463, 3.5417, 3.0355], + device='cuda:1'), covar=tensor([0.1137, 0.1849, 0.0481, 0.2207, 0.1383, 0.0270, 0.0657, 0.0505], + device='cuda:1'), in_proj_covar=tensor([0.0314, 0.0330, 0.0299, 0.0327, 0.0331, 0.0282, 0.0451, 0.0314], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 01:58:05,227 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.341e+02 2.443e+02 3.095e+02 3.828e+02 9.615e+02, threshold=6.189e+02, percent-clipped=4.0 +2023-02-09 01:58:07,399 INFO [train.py:901] (1/4) Epoch 29, batch 2900, loss[loss=0.1665, simple_loss=0.2569, pruned_loss=0.03807, over 8025.00 frames. ], tot_loss[loss=0.197, simple_loss=0.281, pruned_loss=0.05654, over 1614220.68 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:58:21,867 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-02-09 01:58:23,780 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.1435, 1.2844, 1.5705, 1.2765, 0.7918, 1.3494, 1.2107, 0.9914], + device='cuda:1'), covar=tensor([0.0653, 0.1275, 0.1679, 0.1498, 0.0576, 0.1490, 0.0731, 0.0772], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0155, 0.0191, 0.0163, 0.0103, 0.0165, 0.0114, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 01:58:44,210 INFO [train.py:901] (1/4) Epoch 29, batch 2950, loss[loss=0.2074, simple_loss=0.2998, pruned_loss=0.05756, over 8515.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2806, pruned_loss=0.05596, over 1613384.72 frames. ], batch size: 28, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:58:47,081 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-02-09 01:59:02,251 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229297.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:59:17,273 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.610e+02 2.481e+02 3.005e+02 3.741e+02 9.617e+02, threshold=6.010e+02, percent-clipped=3.0 +2023-02-09 01:59:17,399 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229318.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:59:19,290 INFO [train.py:901] (1/4) Epoch 29, batch 3000, loss[loss=0.1786, simple_loss=0.2722, pruned_loss=0.04248, over 8468.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2808, pruned_loss=0.05658, over 1611359.54 frames. ], batch size: 25, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:59:19,290 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 01:59:34,614 INFO [train.py:935] (1/4) Epoch 29, validation: loss=0.17, simple_loss=0.2699, pruned_loss=0.03504, over 944034.00 frames. +2023-02-09 01:59:34,616 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 02:00:09,410 INFO [train.py:901] (1/4) Epoch 29, batch 3050, loss[loss=0.1704, simple_loss=0.266, pruned_loss=0.03739, over 8134.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2811, pruned_loss=0.05644, over 1611296.15 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:00:28,258 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.4187, 4.4203, 3.9473, 2.2787, 3.8937, 4.0025, 3.8282, 3.8091], + device='cuda:1'), covar=tensor([0.0717, 0.0490, 0.0933, 0.4382, 0.0872, 0.1074, 0.1332, 0.0929], + device='cuda:1'), in_proj_covar=tensor([0.0551, 0.0465, 0.0458, 0.0567, 0.0446, 0.0471, 0.0446, 0.0416], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:00:40,614 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229412.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:00:41,971 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9955, 1.9933, 1.8018, 2.6358, 1.4056, 1.7158, 2.0704, 2.1505], + device='cuda:1'), covar=tensor([0.0724, 0.0915, 0.0841, 0.0462, 0.1073, 0.1241, 0.0730, 0.0747], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0196, 0.0244, 0.0213, 0.0203, 0.0247, 0.0250, 0.0205], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 02:00:44,471 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.402e+02 2.505e+02 2.815e+02 3.570e+02 7.212e+02, threshold=5.630e+02, percent-clipped=4.0 +2023-02-09 02:00:46,514 INFO [train.py:901] (1/4) Epoch 29, batch 3100, loss[loss=0.2259, simple_loss=0.3117, pruned_loss=0.07007, over 8294.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2815, pruned_loss=0.05647, over 1613779.56 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:00:59,208 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4597, 2.0721, 3.1591, 1.6875, 1.7747, 3.1234, 0.8445, 2.1383], + device='cuda:1'), covar=tensor([0.1264, 0.1227, 0.0279, 0.1566, 0.2088, 0.0462, 0.2048, 0.1282], + device='cuda:1'), in_proj_covar=tensor([0.0203, 0.0208, 0.0138, 0.0224, 0.0280, 0.0149, 0.0173, 0.0201], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 02:01:15,193 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229461.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:01:21,911 INFO [train.py:901] (1/4) Epoch 29, batch 3150, loss[loss=0.174, simple_loss=0.2639, pruned_loss=0.0421, over 8569.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.282, pruned_loss=0.05712, over 1612943.23 frames. ], batch size: 39, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:01:39,329 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229495.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:01:56,480 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.645e+02 2.412e+02 3.144e+02 3.923e+02 1.015e+03, threshold=6.289e+02, percent-clipped=11.0 +2023-02-09 02:01:58,581 INFO [train.py:901] (1/4) Epoch 29, batch 3200, loss[loss=0.2115, simple_loss=0.2967, pruned_loss=0.06319, over 8341.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.282, pruned_loss=0.05704, over 1612894.32 frames. ], batch size: 26, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:01:58,993 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.49 vs. limit=2.0 +2023-02-09 02:02:02,520 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.72 vs. limit=2.0 +2023-02-09 02:02:14,560 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.91 vs. limit=5.0 +2023-02-09 02:02:29,961 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2889, 2.9121, 2.3172, 2.6015, 2.5834, 2.2206, 2.5377, 2.7717], + device='cuda:1'), covar=tensor([0.1197, 0.0423, 0.0972, 0.0591, 0.0625, 0.1210, 0.0777, 0.0841], + device='cuda:1'), in_proj_covar=tensor([0.0353, 0.0243, 0.0342, 0.0314, 0.0300, 0.0349, 0.0349, 0.0319], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 02:02:35,461 INFO [train.py:901] (1/4) Epoch 29, batch 3250, loss[loss=0.198, simple_loss=0.2846, pruned_loss=0.0557, over 8698.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.282, pruned_loss=0.05669, over 1617897.96 frames. ], batch size: 39, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:02:52,619 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229595.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:03:09,574 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.589e+02 2.257e+02 2.746e+02 3.374e+02 9.131e+02, threshold=5.492e+02, percent-clipped=1.0 +2023-02-09 02:03:11,692 INFO [train.py:901] (1/4) Epoch 29, batch 3300, loss[loss=0.188, simple_loss=0.2814, pruned_loss=0.04735, over 7433.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2817, pruned_loss=0.05601, over 1620572.92 frames. ], batch size: 17, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:03:41,136 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229662.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:03:45,540 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229668.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:03:47,460 INFO [train.py:901] (1/4) Epoch 29, batch 3350, loss[loss=0.2108, simple_loss=0.2842, pruned_loss=0.06866, over 7718.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2823, pruned_loss=0.0567, over 1619262.90 frames. ], batch size: 18, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:03:52,160 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229677.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:04:03,325 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229693.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:04:20,887 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.769e+02 2.561e+02 2.994e+02 3.779e+02 7.703e+02, threshold=5.989e+02, percent-clipped=7.0 +2023-02-09 02:04:22,322 INFO [train.py:901] (1/4) Epoch 29, batch 3400, loss[loss=0.187, simple_loss=0.2581, pruned_loss=0.05793, over 8122.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2822, pruned_loss=0.05637, over 1618465.87 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:04:25,974 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1538, 2.2375, 1.8358, 2.9398, 1.4918, 1.7435, 2.2510, 2.2137], + device='cuda:1'), covar=tensor([0.0683, 0.0785, 0.0894, 0.0308, 0.0933, 0.1182, 0.0734, 0.0847], + device='cuda:1'), in_proj_covar=tensor([0.0230, 0.0195, 0.0245, 0.0213, 0.0203, 0.0246, 0.0251, 0.0205], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 02:04:59,074 INFO [train.py:901] (1/4) Epoch 29, batch 3450, loss[loss=0.193, simple_loss=0.2825, pruned_loss=0.05174, over 8188.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.281, pruned_loss=0.0562, over 1612769.32 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:05:03,510 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229777.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:05:23,715 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229805.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:05:33,258 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.613e+02 2.314e+02 2.735e+02 3.470e+02 1.051e+03, threshold=5.470e+02, percent-clipped=3.0 +2023-02-09 02:05:34,610 INFO [train.py:901] (1/4) Epoch 29, batch 3500, loss[loss=0.2356, simple_loss=0.3191, pruned_loss=0.07605, over 8558.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2805, pruned_loss=0.05585, over 1612654.80 frames. ], batch size: 39, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:05:46,928 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229839.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:05:58,492 WARNING [train.py:1067] (1/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-02-09 02:06:12,017 INFO [train.py:901] (1/4) Epoch 29, batch 3550, loss[loss=0.1873, simple_loss=0.2773, pruned_loss=0.04867, over 8317.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2815, pruned_loss=0.05616, over 1615253.85 frames. ], batch size: 25, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:06:41,730 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229912.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:06:42,432 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.0550, 1.8335, 2.2201, 1.9161, 1.1106, 1.8718, 2.2559, 2.4141], + device='cuda:1'), covar=tensor([0.0441, 0.1243, 0.1550, 0.1334, 0.0594, 0.1435, 0.0637, 0.0542], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0154, 0.0190, 0.0162, 0.0102, 0.0164, 0.0114, 0.0147], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 02:06:44,119 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-09 02:06:46,515 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.708e+02 2.466e+02 3.015e+02 3.666e+02 8.686e+02, threshold=6.030e+02, percent-clipped=2.0 +2023-02-09 02:06:47,434 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229920.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:06:47,947 INFO [train.py:901] (1/4) Epoch 29, batch 3600, loss[loss=0.2175, simple_loss=0.3028, pruned_loss=0.06616, over 8188.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2819, pruned_loss=0.05649, over 1615351.27 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:07:00,990 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229939.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:07:11,753 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229954.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:07:24,202 INFO [train.py:901] (1/4) Epoch 29, batch 3650, loss[loss=0.2507, simple_loss=0.3316, pruned_loss=0.08489, over 8328.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.281, pruned_loss=0.05589, over 1614340.76 frames. ], batch size: 26, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:07:50,393 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230005.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:00,964 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.417e+02 2.398e+02 2.862e+02 3.372e+02 7.881e+02, threshold=5.724e+02, percent-clipped=3.0 +2023-02-09 02:08:01,713 INFO [train.py:901] (1/4) Epoch 29, batch 3700, loss[loss=0.2148, simple_loss=0.2937, pruned_loss=0.06798, over 8341.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2809, pruned_loss=0.05615, over 1615693.30 frames. ], batch size: 49, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:08:02,483 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230021.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:08:06,540 WARNING [train.py:1067] (1/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-02-09 02:08:10,863 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230033.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:25,157 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230054.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:28,463 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230058.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:37,282 INFO [train.py:901] (1/4) Epoch 29, batch 3750, loss[loss=0.1865, simple_loss=0.2766, pruned_loss=0.04822, over 8087.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2823, pruned_loss=0.05678, over 1616854.22 frames. ], batch size: 21, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:08:38,299 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-02-09 02:09:09,097 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.48 vs. limit=5.0 +2023-02-09 02:09:13,042 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.639e+02 2.469e+02 3.110e+02 3.966e+02 1.066e+03, threshold=6.219e+02, percent-clipped=4.0 +2023-02-09 02:09:13,774 INFO [train.py:901] (1/4) Epoch 29, batch 3800, loss[loss=0.1761, simple_loss=0.2545, pruned_loss=0.04883, over 7944.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2826, pruned_loss=0.05716, over 1619841.73 frames. ], batch size: 20, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:09:24,881 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230136.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:09:47,381 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230168.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:09:49,388 INFO [train.py:901] (1/4) Epoch 29, batch 3850, loss[loss=0.1866, simple_loss=0.281, pruned_loss=0.04613, over 8475.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2834, pruned_loss=0.05731, over 1623795.25 frames. ], batch size: 25, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:09:52,571 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-02-09 02:09:53,092 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230176.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:10,985 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230201.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:12,900 WARNING [train.py:1067] (1/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-02-09 02:10:17,569 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230210.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:24,395 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.322e+02 2.511e+02 3.297e+02 3.973e+02 1.066e+03, threshold=6.594e+02, percent-clipped=6.0 +2023-02-09 02:10:25,130 INFO [train.py:901] (1/4) Epoch 29, batch 3900, loss[loss=0.2203, simple_loss=0.2871, pruned_loss=0.07678, over 7660.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2809, pruned_loss=0.05623, over 1618036.47 frames. ], batch size: 19, lr: 2.59e-03, grad_scale: 4.0 +2023-02-09 02:10:36,221 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230235.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:49,413 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-02-09 02:10:51,858 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230256.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:54,122 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230259.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:11:02,669 INFO [train.py:901] (1/4) Epoch 29, batch 3950, loss[loss=0.169, simple_loss=0.241, pruned_loss=0.04849, over 7700.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2809, pruned_loss=0.05632, over 1615711.45 frames. ], batch size: 18, lr: 2.59e-03, grad_scale: 4.0 +2023-02-09 02:11:30,509 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230310.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:11:37,968 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.749e+02 2.564e+02 3.192e+02 4.107e+02 9.729e+02, threshold=6.384e+02, percent-clipped=2.0 +2023-02-09 02:11:38,728 INFO [train.py:901] (1/4) Epoch 29, batch 4000, loss[loss=0.2399, simple_loss=0.3144, pruned_loss=0.08275, over 7058.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2807, pruned_loss=0.05638, over 1609723.44 frames. ], batch size: 71, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:11:49,754 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230335.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:11:59,723 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230349.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:12:16,079 INFO [train.py:901] (1/4) Epoch 29, batch 4050, loss[loss=0.2085, simple_loss=0.2895, pruned_loss=0.0638, over 8100.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2815, pruned_loss=0.05698, over 1613552.41 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:12:16,261 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230371.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:12:31,112 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230392.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:12:43,589 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2386, 1.2719, 1.2345, 1.9471, 0.8451, 1.1183, 1.5902, 1.4333], + device='cuda:1'), covar=tensor([0.1495, 0.1124, 0.1906, 0.0489, 0.1242, 0.1965, 0.0690, 0.0863], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0196, 0.0244, 0.0213, 0.0202, 0.0247, 0.0251, 0.0204], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 02:12:48,555 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230417.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:12:50,324 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.613e+02 2.354e+02 2.823e+02 3.644e+02 9.834e+02, threshold=5.645e+02, percent-clipped=2.0 +2023-02-09 02:12:51,013 INFO [train.py:901] (1/4) Epoch 29, batch 4100, loss[loss=0.1938, simple_loss=0.2827, pruned_loss=0.05244, over 8802.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2815, pruned_loss=0.05679, over 1613697.09 frames. ], batch size: 49, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:13:22,439 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230464.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:13:26,984 INFO [train.py:901] (1/4) Epoch 29, batch 4150, loss[loss=0.2359, simple_loss=0.3177, pruned_loss=0.07708, over 8650.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2833, pruned_loss=0.05775, over 1612561.22 frames. ], batch size: 27, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:13:33,237 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-02-09 02:13:38,467 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230486.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:13:46,435 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=4.29 vs. limit=5.0 +2023-02-09 02:13:56,255 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230512.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:14:01,408 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1689, 1.9537, 2.4747, 2.0683, 2.4454, 2.2445, 2.1151, 1.4329], + device='cuda:1'), covar=tensor([0.5951, 0.5439, 0.2193, 0.4151, 0.2740, 0.3462, 0.1988, 0.5703], + device='cuda:1'), in_proj_covar=tensor([0.0968, 0.1036, 0.0841, 0.1007, 0.1029, 0.0941, 0.0778, 0.0857], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:14:01,810 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.657e+02 2.471e+02 3.120e+02 4.398e+02 1.014e+03, threshold=6.241e+02, percent-clipped=11.0 +2023-02-09 02:14:02,515 INFO [train.py:901] (1/4) Epoch 29, batch 4200, loss[loss=0.2178, simple_loss=0.3069, pruned_loss=0.06432, over 8317.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2829, pruned_loss=0.05737, over 1612366.46 frames. ], batch size: 25, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:14:16,937 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-02-09 02:14:17,784 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230543.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:14:24,144 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230551.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:14:38,390 INFO [train.py:901] (1/4) Epoch 29, batch 4250, loss[loss=0.1862, simple_loss=0.2717, pruned_loss=0.05037, over 7934.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2822, pruned_loss=0.05704, over 1610183.21 frames. ], batch size: 20, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:14:41,195 WARNING [train.py:1067] (1/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-02-09 02:15:01,587 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230603.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:13,975 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.550e+02 2.532e+02 3.108e+02 3.832e+02 7.900e+02, threshold=6.217e+02, percent-clipped=4.0 +2023-02-09 02:15:14,735 INFO [train.py:901] (1/4) Epoch 29, batch 4300, loss[loss=0.1682, simple_loss=0.2504, pruned_loss=0.04302, over 8084.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2827, pruned_loss=0.05766, over 1614359.10 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:15:19,483 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230627.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:19,525 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230627.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:37,548 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230652.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:51,725 INFO [train.py:901] (1/4) Epoch 29, batch 4350, loss[loss=0.1944, simple_loss=0.2791, pruned_loss=0.05488, over 7984.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2833, pruned_loss=0.058, over 1616629.27 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:15:55,694 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2363, 1.1066, 1.3214, 1.0111, 1.0946, 1.3502, 0.0950, 0.8983], + device='cuda:1'), covar=tensor([0.1454, 0.1222, 0.0501, 0.0625, 0.2179, 0.0531, 0.1899, 0.1188], + device='cuda:1'), in_proj_covar=tensor([0.0203, 0.0206, 0.0138, 0.0225, 0.0281, 0.0149, 0.0174, 0.0202], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 02:16:17,530 WARNING [train.py:1067] (1/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-02-09 02:16:27,817 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230718.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:16:29,089 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.602e+02 2.462e+02 3.042e+02 3.743e+02 1.027e+03, threshold=6.085e+02, percent-clipped=1.0 +2023-02-09 02:16:29,336 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230720.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:16:29,846 INFO [train.py:901] (1/4) Epoch 29, batch 4400, loss[loss=0.2496, simple_loss=0.3174, pruned_loss=0.09086, over 8483.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2829, pruned_loss=0.05782, over 1618389.51 frames. ], batch size: 27, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:16:47,635 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230745.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:16:57,339 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-02-09 02:17:06,602 INFO [train.py:901] (1/4) Epoch 29, batch 4450, loss[loss=0.1677, simple_loss=0.2555, pruned_loss=0.03995, over 7813.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2826, pruned_loss=0.05759, over 1620241.95 frames. ], batch size: 20, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:17:36,256 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230810.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:17:43,245 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.628e+02 2.373e+02 2.913e+02 3.542e+02 8.795e+02, threshold=5.826e+02, percent-clipped=1.0 +2023-02-09 02:17:43,983 INFO [train.py:901] (1/4) Epoch 29, batch 4500, loss[loss=0.2174, simple_loss=0.3099, pruned_loss=0.06242, over 8277.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2828, pruned_loss=0.05747, over 1618499.05 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:17:49,304 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8609, 1.5128, 3.4274, 1.6919, 2.3500, 3.7100, 3.8075, 3.2027], + device='cuda:1'), covar=tensor([0.1298, 0.1917, 0.0323, 0.1998, 0.1098, 0.0238, 0.0600, 0.0539], + device='cuda:1'), in_proj_covar=tensor([0.0311, 0.0328, 0.0296, 0.0325, 0.0328, 0.0279, 0.0447, 0.0311], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 02:17:51,400 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230830.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:17:54,895 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-02-09 02:18:19,714 INFO [train.py:901] (1/4) Epoch 29, batch 4550, loss[loss=0.1852, simple_loss=0.2686, pruned_loss=0.05087, over 7417.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2829, pruned_loss=0.05743, over 1615920.66 frames. ], batch size: 17, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:18:28,415 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230883.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:31,629 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230887.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:37,202 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230895.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:46,427 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230908.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:46,490 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230908.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:55,066 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.629e+02 2.340e+02 2.935e+02 3.730e+02 9.176e+02, threshold=5.869e+02, percent-clipped=5.0 +2023-02-09 02:18:55,765 INFO [train.py:901] (1/4) Epoch 29, batch 4600, loss[loss=0.2127, simple_loss=0.3126, pruned_loss=0.0564, over 8329.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2829, pruned_loss=0.05783, over 1613049.09 frames. ], batch size: 25, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:19:13,588 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230945.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:31,779 INFO [train.py:901] (1/4) Epoch 29, batch 4650, loss[loss=0.1994, simple_loss=0.2949, pruned_loss=0.05197, over 8106.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2827, pruned_loss=0.05776, over 1614951.87 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:19:33,907 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230974.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:42,737 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.4524, 4.4002, 3.9735, 2.0332, 3.9098, 4.1147, 3.9414, 3.9547], + device='cuda:1'), covar=tensor([0.0694, 0.0545, 0.0964, 0.4519, 0.0901, 0.0853, 0.1315, 0.0870], + device='cuda:1'), in_proj_covar=tensor([0.0544, 0.0460, 0.0451, 0.0563, 0.0442, 0.0466, 0.0441, 0.0411], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:19:51,178 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230999.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:53,286 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231002.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:59,427 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231010.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:20:06,387 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.812e+02 2.522e+02 3.157e+02 3.845e+02 7.559e+02, threshold=6.314e+02, percent-clipped=7.0 +2023-02-09 02:20:07,122 INFO [train.py:901] (1/4) Epoch 29, batch 4700, loss[loss=0.1976, simple_loss=0.2867, pruned_loss=0.05426, over 8092.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2823, pruned_loss=0.05729, over 1616068.07 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:20:43,864 INFO [train.py:901] (1/4) Epoch 29, batch 4750, loss[loss=0.1963, simple_loss=0.2889, pruned_loss=0.05189, over 8451.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2827, pruned_loss=0.05699, over 1621400.49 frames. ], batch size: 27, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:21:00,606 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-02-09 02:21:02,784 WARNING [train.py:1067] (1/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-02-09 02:21:11,853 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5110, 2.3386, 1.7258, 2.1499, 2.0215, 1.4870, 1.9866, 1.9673], + device='cuda:1'), covar=tensor([0.1333, 0.0447, 0.1309, 0.0559, 0.0723, 0.1561, 0.0913, 0.0846], + device='cuda:1'), in_proj_covar=tensor([0.0356, 0.0244, 0.0344, 0.0314, 0.0301, 0.0348, 0.0351, 0.0319], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 02:21:18,720 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.727e+02 2.440e+02 2.924e+02 3.615e+02 6.392e+02, threshold=5.847e+02, percent-clipped=1.0 +2023-02-09 02:21:19,439 INFO [train.py:901] (1/4) Epoch 29, batch 4800, loss[loss=0.1453, simple_loss=0.2346, pruned_loss=0.02797, over 8080.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2831, pruned_loss=0.05754, over 1622920.31 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:21:26,337 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7634, 2.3876, 4.1175, 1.7763, 3.1490, 2.4482, 1.8536, 3.0388], + device='cuda:1'), covar=tensor([0.2024, 0.2767, 0.0877, 0.4727, 0.1789, 0.3296, 0.2559, 0.2408], + device='cuda:1'), in_proj_covar=tensor([0.0543, 0.0644, 0.0567, 0.0678, 0.0668, 0.0616, 0.0570, 0.0649], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:21:43,702 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=231154.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:21:43,876 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3556, 2.3181, 2.9721, 2.4108, 2.9532, 2.5652, 2.4031, 1.9038], + device='cuda:1'), covar=tensor([0.6014, 0.5306, 0.2312, 0.4349, 0.2864, 0.3191, 0.1902, 0.5901], + device='cuda:1'), in_proj_covar=tensor([0.0970, 0.1035, 0.0841, 0.1007, 0.1028, 0.0940, 0.0779, 0.0858], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:21:55,947 INFO [train.py:901] (1/4) Epoch 29, batch 4850, loss[loss=0.16, simple_loss=0.244, pruned_loss=0.03795, over 7255.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2819, pruned_loss=0.05669, over 1623913.69 frames. ], batch size: 16, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:21:55,956 WARNING [train.py:1067] (1/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-02-09 02:22:17,915 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231201.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:22:31,032 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.570e+02 2.490e+02 3.133e+02 4.186e+02 8.287e+02, threshold=6.266e+02, percent-clipped=6.0 +2023-02-09 02:22:31,778 INFO [train.py:901] (1/4) Epoch 29, batch 4900, loss[loss=0.1656, simple_loss=0.2491, pruned_loss=0.04104, over 7785.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2823, pruned_loss=0.05691, over 1618903.91 frames. ], batch size: 19, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:22:35,623 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231226.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:22:53,270 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.82 vs. limit=5.0 +2023-02-09 02:22:54,381 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=231252.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:22:59,238 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231258.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:04,600 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231266.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:06,606 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=231269.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:06,641 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231269.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:07,928 INFO [train.py:901] (1/4) Epoch 29, batch 4950, loss[loss=0.2069, simple_loss=0.2956, pruned_loss=0.05904, over 8608.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2835, pruned_loss=0.05765, over 1618958.61 frames. ], batch size: 34, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:23:17,261 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231283.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:22,755 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231291.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:43,160 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.636e+02 2.476e+02 2.910e+02 3.581e+02 7.956e+02, threshold=5.820e+02, percent-clipped=2.0 +2023-02-09 02:23:43,853 INFO [train.py:901] (1/4) Epoch 29, batch 5000, loss[loss=0.2223, simple_loss=0.3109, pruned_loss=0.06691, over 8608.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2824, pruned_loss=0.0572, over 1618724.65 frames. ], batch size: 49, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:23:45,410 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5184, 1.2902, 2.2214, 1.2852, 2.2205, 2.3914, 2.5369, 2.0270], + device='cuda:1'), covar=tensor([0.1032, 0.1401, 0.0467, 0.1959, 0.0809, 0.0413, 0.0834, 0.0684], + device='cuda:1'), in_proj_covar=tensor([0.0313, 0.0330, 0.0299, 0.0327, 0.0328, 0.0281, 0.0450, 0.0312], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 02:24:17,059 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231367.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:24:19,691 INFO [train.py:901] (1/4) Epoch 29, batch 5050, loss[loss=0.2384, simple_loss=0.3168, pruned_loss=0.07997, over 8545.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2814, pruned_loss=0.05671, over 1620821.60 frames. ], batch size: 49, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:24:37,781 WARNING [train.py:1067] (1/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-02-09 02:24:55,715 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.315e+02 2.860e+02 3.491e+02 8.708e+02, threshold=5.721e+02, percent-clipped=8.0 +2023-02-09 02:24:56,368 INFO [train.py:901] (1/4) Epoch 29, batch 5100, loss[loss=0.1817, simple_loss=0.2513, pruned_loss=0.05603, over 7443.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2806, pruned_loss=0.05663, over 1619078.84 frames. ], batch size: 17, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:25:32,299 INFO [train.py:901] (1/4) Epoch 29, batch 5150, loss[loss=0.2292, simple_loss=0.3197, pruned_loss=0.0693, over 8142.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2805, pruned_loss=0.05648, over 1619096.59 frames. ], batch size: 22, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:25:52,005 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-02-09 02:26:07,335 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-02-09 02:26:08,155 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.585e+02 2.313e+02 2.816e+02 3.592e+02 7.666e+02, threshold=5.632e+02, percent-clipped=6.0 +2023-02-09 02:26:08,940 INFO [train.py:901] (1/4) Epoch 29, batch 5200, loss[loss=0.1977, simple_loss=0.2909, pruned_loss=0.05229, over 8248.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2808, pruned_loss=0.05662, over 1618155.92 frames. ], batch size: 24, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:26:12,095 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231525.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:26:30,200 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231550.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:26:39,125 WARNING [train.py:1067] (1/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-02-09 02:26:44,754 INFO [train.py:901] (1/4) Epoch 29, batch 5250, loss[loss=0.2286, simple_loss=0.3081, pruned_loss=0.07453, over 8508.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2805, pruned_loss=0.05661, over 1614795.43 frames. ], batch size: 28, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:27:15,796 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=231613.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:20,503 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.520e+02 2.485e+02 2.947e+02 3.559e+02 7.815e+02, threshold=5.893e+02, percent-clipped=2.0 +2023-02-09 02:27:21,229 INFO [train.py:901] (1/4) Epoch 29, batch 5300, loss[loss=0.1937, simple_loss=0.2774, pruned_loss=0.05505, over 8122.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2806, pruned_loss=0.05654, over 1613525.46 frames. ], batch size: 22, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:27:22,849 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231623.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:41,043 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231648.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:46,366 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.1019, 1.5176, 4.2855, 1.7818, 3.8045, 3.5911, 3.8826, 3.7929], + device='cuda:1'), covar=tensor([0.0711, 0.4739, 0.0599, 0.4157, 0.1098, 0.1000, 0.0625, 0.0704], + device='cuda:1'), in_proj_covar=tensor([0.0682, 0.0660, 0.0742, 0.0653, 0.0739, 0.0632, 0.0640, 0.0717], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:27:47,752 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4444, 2.6305, 2.2468, 3.9822, 1.6662, 2.0165, 2.4005, 2.6487], + device='cuda:1'), covar=tensor([0.0701, 0.0846, 0.0785, 0.0243, 0.1060, 0.1237, 0.0958, 0.0808], + device='cuda:1'), in_proj_covar=tensor([0.0230, 0.0195, 0.0243, 0.0213, 0.0202, 0.0245, 0.0250, 0.0204], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 02:27:53,030 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=231665.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:54,362 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.1421, 1.5280, 4.3469, 1.5999, 3.8276, 3.6086, 3.9179, 3.8180], + device='cuda:1'), covar=tensor([0.0697, 0.4473, 0.0558, 0.4373, 0.1105, 0.1005, 0.0664, 0.0724], + device='cuda:1'), in_proj_covar=tensor([0.0682, 0.0660, 0.0742, 0.0653, 0.0739, 0.0632, 0.0640, 0.0717], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:27:57,008 INFO [train.py:901] (1/4) Epoch 29, batch 5350, loss[loss=0.2253, simple_loss=0.303, pruned_loss=0.07375, over 8490.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.281, pruned_loss=0.0566, over 1609454.88 frames. ], batch size: 39, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:28:32,684 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.493e+02 2.535e+02 3.138e+02 3.956e+02 6.651e+02, threshold=6.276e+02, percent-clipped=5.0 +2023-02-09 02:28:33,431 INFO [train.py:901] (1/4) Epoch 29, batch 5400, loss[loss=0.1947, simple_loss=0.2867, pruned_loss=0.05136, over 8667.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2811, pruned_loss=0.05701, over 1610618.20 frames. ], batch size: 34, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:28:38,406 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231728.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:29:09,675 INFO [train.py:901] (1/4) Epoch 29, batch 5450, loss[loss=0.2381, simple_loss=0.3165, pruned_loss=0.07981, over 8501.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2796, pruned_loss=0.05621, over 1608447.62 frames. ], batch size: 26, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:29:24,004 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4943, 2.3222, 1.7413, 2.2055, 1.9091, 1.4661, 1.9577, 1.9472], + device='cuda:1'), covar=tensor([0.1374, 0.0506, 0.1298, 0.0612, 0.0917, 0.1744, 0.1005, 0.0977], + device='cuda:1'), in_proj_covar=tensor([0.0356, 0.0244, 0.0343, 0.0315, 0.0302, 0.0348, 0.0351, 0.0320], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 02:29:29,513 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-02-09 02:29:43,731 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7087, 1.9612, 2.0345, 1.3549, 2.2141, 1.4849, 0.7296, 1.9187], + device='cuda:1'), covar=tensor([0.0761, 0.0448, 0.0348, 0.0771, 0.0508, 0.1034, 0.1120, 0.0445], + device='cuda:1'), in_proj_covar=tensor([0.0476, 0.0417, 0.0371, 0.0465, 0.0400, 0.0556, 0.0408, 0.0445], + device='cuda:1'), out_proj_covar=tensor([1.2593e-04, 1.0787e-04, 9.6390e-05, 1.2157e-04, 1.0474e-04, 1.5474e-04, + 1.0873e-04, 1.1633e-04], device='cuda:1') +2023-02-09 02:29:44,878 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.390e+02 2.416e+02 2.826e+02 3.521e+02 6.915e+02, threshold=5.653e+02, percent-clipped=1.0 +2023-02-09 02:29:45,630 INFO [train.py:901] (1/4) Epoch 29, batch 5500, loss[loss=0.1606, simple_loss=0.246, pruned_loss=0.03761, over 7928.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2797, pruned_loss=0.05657, over 1606132.12 frames. ], batch size: 20, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:29:48,659 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8179, 1.4463, 1.8473, 1.3759, 0.8932, 1.5761, 1.5897, 1.5470], + device='cuda:1'), covar=tensor([0.0590, 0.1299, 0.1650, 0.1505, 0.0624, 0.1454, 0.0753, 0.0679], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0155, 0.0191, 0.0163, 0.0102, 0.0165, 0.0114, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 02:30:21,383 INFO [train.py:901] (1/4) Epoch 29, batch 5550, loss[loss=0.2009, simple_loss=0.2822, pruned_loss=0.05978, over 7967.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2799, pruned_loss=0.05667, over 1611209.73 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:30:49,872 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-02-09 02:30:56,571 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.629e+02 2.336e+02 2.917e+02 3.506e+02 1.057e+03, threshold=5.834e+02, percent-clipped=5.0 +2023-02-09 02:30:57,329 INFO [train.py:901] (1/4) Epoch 29, batch 5600, loss[loss=0.2183, simple_loss=0.2938, pruned_loss=0.07139, over 8101.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2804, pruned_loss=0.05698, over 1611387.78 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:31:24,823 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.51 vs. limit=2.0 +2023-02-09 02:31:34,496 INFO [train.py:901] (1/4) Epoch 29, batch 5650, loss[loss=0.1655, simple_loss=0.2509, pruned_loss=0.04008, over 7787.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2798, pruned_loss=0.05663, over 1608139.87 frames. ], batch size: 19, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:31:38,801 WARNING [train.py:1067] (1/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-02-09 02:31:43,800 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231984.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:32:02,828 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232009.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:32:02,957 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=232009.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:32:10,312 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.351e+02 2.719e+02 3.536e+02 6.635e+02, threshold=5.438e+02, percent-clipped=1.0 +2023-02-09 02:32:11,021 INFO [train.py:901] (1/4) Epoch 29, batch 5700, loss[loss=0.2154, simple_loss=0.301, pruned_loss=0.06491, over 8372.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.281, pruned_loss=0.05671, over 1612058.09 frames. ], batch size: 48, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:32:45,188 WARNING [train.py:1067] (1/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-02-09 02:32:47,252 INFO [train.py:901] (1/4) Epoch 29, batch 5750, loss[loss=0.2078, simple_loss=0.3015, pruned_loss=0.05705, over 8479.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2819, pruned_loss=0.05694, over 1616047.65 frames. ], batch size: 25, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:32:54,527 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.9678, 2.4320, 3.6936, 1.8814, 2.0717, 3.6739, 0.6280, 2.1687], + device='cuda:1'), covar=tensor([0.1306, 0.1120, 0.0265, 0.1477, 0.2078, 0.0279, 0.2083, 0.1371], + device='cuda:1'), in_proj_covar=tensor([0.0202, 0.0205, 0.0137, 0.0224, 0.0278, 0.0149, 0.0172, 0.0200], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 02:33:07,927 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7615, 2.3720, 4.1958, 1.7043, 3.1033, 2.3737, 1.9881, 3.0565], + device='cuda:1'), covar=tensor([0.1976, 0.2792, 0.0798, 0.4746, 0.1822, 0.3288, 0.2412, 0.2334], + device='cuda:1'), in_proj_covar=tensor([0.0544, 0.0642, 0.0566, 0.0676, 0.0667, 0.0615, 0.0569, 0.0647], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:33:23,822 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.487e+02 2.417e+02 3.013e+02 3.730e+02 1.097e+03, threshold=6.026e+02, percent-clipped=6.0 +2023-02-09 02:33:24,554 INFO [train.py:901] (1/4) Epoch 29, batch 5800, loss[loss=0.1635, simple_loss=0.2591, pruned_loss=0.03401, over 7805.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.282, pruned_loss=0.05637, over 1618985.95 frames. ], batch size: 20, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:33:26,837 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232124.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:33:38,341 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9566, 1.5531, 3.1182, 1.2422, 2.3358, 3.3790, 3.6511, 2.5642], + device='cuda:1'), covar=tensor([0.1393, 0.2112, 0.0458, 0.2937, 0.1189, 0.0397, 0.0580, 0.0963], + device='cuda:1'), in_proj_covar=tensor([0.0310, 0.0328, 0.0295, 0.0325, 0.0326, 0.0279, 0.0446, 0.0310], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 02:33:46,044 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7270, 2.0172, 2.0971, 1.3719, 2.1897, 1.5377, 0.6943, 2.0128], + device='cuda:1'), covar=tensor([0.0694, 0.0412, 0.0337, 0.0706, 0.0452, 0.1061, 0.1044, 0.0383], + device='cuda:1'), in_proj_covar=tensor([0.0477, 0.0417, 0.0371, 0.0465, 0.0400, 0.0558, 0.0407, 0.0445], + device='cuda:1'), out_proj_covar=tensor([1.2632e-04, 1.0779e-04, 9.6523e-05, 1.2158e-04, 1.0470e-04, 1.5528e-04, + 1.0856e-04, 1.1647e-04], device='cuda:1') +2023-02-09 02:33:59,562 INFO [train.py:901] (1/4) Epoch 29, batch 5850, loss[loss=0.1607, simple_loss=0.2504, pruned_loss=0.03545, over 7917.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2806, pruned_loss=0.05555, over 1618849.16 frames. ], batch size: 20, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:34:34,659 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.378e+02 2.421e+02 2.869e+02 3.503e+02 7.290e+02, threshold=5.737e+02, percent-clipped=3.0 +2023-02-09 02:34:35,359 INFO [train.py:901] (1/4) Epoch 29, batch 5900, loss[loss=0.1946, simple_loss=0.2802, pruned_loss=0.05456, over 8579.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2805, pruned_loss=0.05587, over 1617900.65 frames. ], batch size: 31, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:34:41,344 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.96 vs. limit=5.0 +2023-02-09 02:34:48,937 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-02-09 02:35:06,627 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232265.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:35:10,677 INFO [train.py:901] (1/4) Epoch 29, batch 5950, loss[loss=0.2386, simple_loss=0.3183, pruned_loss=0.07947, over 8696.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2807, pruned_loss=0.05629, over 1614119.03 frames. ], batch size: 34, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:35:13,134 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.48 vs. limit=2.0 +2023-02-09 02:35:18,589 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2459, 2.0533, 2.6541, 2.2403, 2.6238, 2.3249, 2.1597, 1.6032], + device='cuda:1'), covar=tensor([0.5741, 0.4924, 0.2065, 0.3643, 0.2452, 0.2974, 0.1845, 0.5137], + device='cuda:1'), in_proj_covar=tensor([0.0978, 0.1039, 0.0845, 0.1010, 0.1033, 0.0942, 0.0781, 0.0860], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:35:46,600 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.609e+02 2.503e+02 2.837e+02 3.700e+02 9.228e+02, threshold=5.675e+02, percent-clipped=4.0 +2023-02-09 02:35:47,347 INFO [train.py:901] (1/4) Epoch 29, batch 6000, loss[loss=0.1991, simple_loss=0.2878, pruned_loss=0.05516, over 7976.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2805, pruned_loss=0.05659, over 1612740.05 frames. ], batch size: 21, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:35:47,348 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 02:35:55,517 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.8517, 3.8202, 3.5342, 2.0041, 3.3654, 3.5613, 3.3866, 3.4506], + device='cuda:1'), covar=tensor([0.0781, 0.0456, 0.0801, 0.4722, 0.0927, 0.0802, 0.1201, 0.0827], + device='cuda:1'), in_proj_covar=tensor([0.0543, 0.0459, 0.0448, 0.0560, 0.0441, 0.0467, 0.0441, 0.0410], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:36:01,202 INFO [train.py:935] (1/4) Epoch 29, validation: loss=0.1708, simple_loss=0.2701, pruned_loss=0.03577, over 944034.00 frames. +2023-02-09 02:36:01,203 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 02:36:15,989 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.2381, 3.1057, 2.9208, 1.5526, 2.8526, 2.9256, 2.8683, 2.7649], + device='cuda:1'), covar=tensor([0.1266, 0.0921, 0.1470, 0.4689, 0.1203, 0.1288, 0.1765, 0.1188], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0460, 0.0450, 0.0562, 0.0442, 0.0469, 0.0442, 0.0412], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:36:37,559 INFO [train.py:901] (1/4) Epoch 29, batch 6050, loss[loss=0.1804, simple_loss=0.2644, pruned_loss=0.04814, over 8098.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2795, pruned_loss=0.05615, over 1610569.11 frames. ], batch size: 21, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:36:44,009 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=232380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:36:57,843 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232399.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:37:02,394 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=232405.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:37:12,730 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.752e+02 2.608e+02 3.073e+02 4.031e+02 7.869e+02, threshold=6.145e+02, percent-clipped=3.0 +2023-02-09 02:37:13,450 INFO [train.py:901] (1/4) Epoch 29, batch 6100, loss[loss=0.1813, simple_loss=0.2714, pruned_loss=0.04562, over 8501.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2793, pruned_loss=0.05602, over 1612706.04 frames. ], batch size: 29, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:37:27,131 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-02-09 02:37:50,002 INFO [train.py:901] (1/4) Epoch 29, batch 6150, loss[loss=0.1854, simple_loss=0.2782, pruned_loss=0.04629, over 8027.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2811, pruned_loss=0.05687, over 1613595.95 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:38:06,444 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232494.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:38:25,292 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.683e+02 2.548e+02 2.901e+02 3.636e+02 6.365e+02, threshold=5.801e+02, percent-clipped=1.0 +2023-02-09 02:38:25,864 INFO [train.py:901] (1/4) Epoch 29, batch 6200, loss[loss=0.2513, simple_loss=0.2941, pruned_loss=0.1042, over 7545.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2812, pruned_loss=0.05733, over 1609900.51 frames. ], batch size: 18, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:38:36,670 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232536.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:38:48,576 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3802, 2.3067, 1.7615, 2.0910, 1.8869, 1.5306, 1.8283, 1.8634], + device='cuda:1'), covar=tensor([0.1675, 0.0547, 0.1389, 0.0728, 0.0887, 0.1726, 0.1115, 0.1036], + device='cuda:1'), in_proj_covar=tensor([0.0358, 0.0245, 0.0345, 0.0315, 0.0303, 0.0351, 0.0352, 0.0321], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 02:38:53,536 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0387, 1.8769, 2.2903, 1.9838, 2.2571, 2.1262, 1.9782, 1.2492], + device='cuda:1'), covar=tensor([0.5694, 0.4925, 0.2150, 0.3733, 0.2632, 0.3429, 0.2065, 0.5424], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1036, 0.0843, 0.1005, 0.1030, 0.0939, 0.0779, 0.0857], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:39:02,789 INFO [train.py:901] (1/4) Epoch 29, batch 6250, loss[loss=0.1875, simple_loss=0.276, pruned_loss=0.04949, over 8505.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2808, pruned_loss=0.05701, over 1611075.19 frames. ], batch size: 29, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:39:07,143 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7176, 1.3933, 1.6855, 1.3020, 0.8677, 1.4555, 1.6166, 1.5169], + device='cuda:1'), covar=tensor([0.0617, 0.1309, 0.1654, 0.1542, 0.0621, 0.1547, 0.0754, 0.0648], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0155, 0.0190, 0.0162, 0.0102, 0.0164, 0.0114, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 02:39:16,229 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1499, 1.7330, 3.4065, 1.6320, 2.5564, 3.7839, 3.8250, 3.2950], + device='cuda:1'), covar=tensor([0.1221, 0.1843, 0.0377, 0.2167, 0.1171, 0.0236, 0.0636, 0.0520], + device='cuda:1'), in_proj_covar=tensor([0.0310, 0.0327, 0.0295, 0.0326, 0.0325, 0.0278, 0.0445, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 02:39:23,932 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8692, 1.3578, 3.3814, 1.6740, 2.5381, 3.6615, 3.7769, 3.1746], + device='cuda:1'), covar=tensor([0.1200, 0.1965, 0.0286, 0.1959, 0.0854, 0.0221, 0.0585, 0.0524], + device='cuda:1'), in_proj_covar=tensor([0.0310, 0.0326, 0.0295, 0.0326, 0.0325, 0.0278, 0.0445, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 02:39:30,146 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232609.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:39:37,831 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.584e+02 2.332e+02 2.762e+02 3.491e+02 8.673e+02, threshold=5.523e+02, percent-clipped=4.0 +2023-02-09 02:39:39,193 INFO [train.py:901] (1/4) Epoch 29, batch 6300, loss[loss=0.1938, simple_loss=0.2737, pruned_loss=0.05702, over 7977.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2799, pruned_loss=0.05623, over 1610700.94 frames. ], batch size: 21, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:39:56,294 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.5243, 2.3571, 3.0775, 2.4237, 2.9755, 2.6019, 2.5561, 1.9150], + device='cuda:1'), covar=tensor([0.5884, 0.5797, 0.2423, 0.4811, 0.3070, 0.3704, 0.1855, 0.6350], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1036, 0.0844, 0.1006, 0.1030, 0.0940, 0.0780, 0.0858], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:40:12,352 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6792, 2.3063, 3.8685, 1.5454, 2.7314, 2.2825, 1.7754, 2.9525], + device='cuda:1'), covar=tensor([0.2089, 0.2859, 0.0862, 0.5037, 0.2223, 0.3414, 0.2755, 0.2489], + device='cuda:1'), in_proj_covar=tensor([0.0544, 0.0641, 0.0565, 0.0674, 0.0667, 0.0614, 0.0569, 0.0648], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:40:14,953 INFO [train.py:901] (1/4) Epoch 29, batch 6350, loss[loss=0.2349, simple_loss=0.315, pruned_loss=0.07737, over 8723.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2795, pruned_loss=0.05571, over 1612201.21 frames. ], batch size: 34, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:40:50,678 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.474e+02 2.465e+02 3.018e+02 3.778e+02 1.284e+03, threshold=6.036e+02, percent-clipped=3.0 +2023-02-09 02:40:51,324 INFO [train.py:901] (1/4) Epoch 29, batch 6400, loss[loss=0.203, simple_loss=0.2967, pruned_loss=0.05469, over 8490.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2804, pruned_loss=0.05638, over 1613833.59 frames. ], batch size: 26, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:40:53,598 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232724.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:41:08,031 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232743.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:41:28,293 INFO [train.py:901] (1/4) Epoch 29, batch 6450, loss[loss=0.2147, simple_loss=0.3077, pruned_loss=0.06083, over 8285.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2806, pruned_loss=0.05617, over 1615831.34 frames. ], batch size: 23, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:42:05,016 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.603e+02 2.358e+02 3.084e+02 4.266e+02 9.590e+02, threshold=6.168e+02, percent-clipped=5.0 +2023-02-09 02:42:05,753 INFO [train.py:901] (1/4) Epoch 29, batch 6500, loss[loss=0.2111, simple_loss=0.2913, pruned_loss=0.06551, over 8777.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2817, pruned_loss=0.05626, over 1625208.79 frames. ], batch size: 32, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:42:17,436 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232838.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:42:24,000 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.78 vs. limit=2.0 +2023-02-09 02:42:24,514 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6111, 2.5591, 1.8500, 2.2811, 2.1940, 1.6029, 2.1214, 2.1680], + device='cuda:1'), covar=tensor([0.1638, 0.0470, 0.1334, 0.0682, 0.0786, 0.1681, 0.1056, 0.1073], + device='cuda:1'), in_proj_covar=tensor([0.0358, 0.0246, 0.0347, 0.0316, 0.0303, 0.0351, 0.0352, 0.0321], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 02:42:31,253 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232858.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:42:31,450 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.78 vs. limit=2.0 +2023-02-09 02:42:40,786 INFO [train.py:901] (1/4) Epoch 29, batch 6550, loss[loss=0.2392, simple_loss=0.3183, pruned_loss=0.08002, over 8624.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2814, pruned_loss=0.05648, over 1621748.51 frames. ], batch size: 34, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:42:47,143 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232880.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:42:47,794 WARNING [train.py:1067] (1/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-02-09 02:43:06,964 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3169, 2.1320, 1.6854, 1.9234, 1.8060, 1.4451, 1.7889, 1.6710], + device='cuda:1'), covar=tensor([0.1414, 0.0459, 0.1292, 0.0624, 0.0767, 0.1688, 0.0969, 0.0960], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0247, 0.0348, 0.0317, 0.0304, 0.0352, 0.0353, 0.0322], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 02:43:08,193 WARNING [train.py:1067] (1/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-02-09 02:43:16,620 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.606e+02 2.398e+02 2.846e+02 3.696e+02 7.042e+02, threshold=5.692e+02, percent-clipped=2.0 +2023-02-09 02:43:17,343 INFO [train.py:901] (1/4) Epoch 29, batch 6600, loss[loss=0.1658, simple_loss=0.2471, pruned_loss=0.04226, over 7528.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2816, pruned_loss=0.05649, over 1620464.15 frames. ], batch size: 18, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:43:40,629 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232953.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:43:52,852 INFO [train.py:901] (1/4) Epoch 29, batch 6650, loss[loss=0.224, simple_loss=0.3094, pruned_loss=0.06934, over 8130.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.281, pruned_loss=0.05626, over 1618967.03 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:43:59,974 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=232980.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:10,260 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232995.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:17,501 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233005.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:19,561 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233008.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:28,483 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.750e+02 2.388e+02 2.820e+02 3.481e+02 6.998e+02, threshold=5.640e+02, percent-clipped=2.0 +2023-02-09 02:44:28,503 INFO [train.py:901] (1/4) Epoch 29, batch 6700, loss[loss=0.2413, simple_loss=0.3242, pruned_loss=0.07917, over 8140.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2805, pruned_loss=0.05644, over 1617560.09 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:44:40,615 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233036.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:55,114 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7863, 2.2937, 4.1267, 1.5822, 2.9034, 2.4044, 1.8503, 3.0555], + device='cuda:1'), covar=tensor([0.2044, 0.2798, 0.0867, 0.4943, 0.2043, 0.3241, 0.2610, 0.2414], + device='cuda:1'), in_proj_covar=tensor([0.0543, 0.0640, 0.0565, 0.0675, 0.0665, 0.0614, 0.0567, 0.0646], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:45:06,272 INFO [train.py:901] (1/4) Epoch 29, batch 6750, loss[loss=0.1809, simple_loss=0.255, pruned_loss=0.05339, over 7434.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2802, pruned_loss=0.05662, over 1613743.00 frames. ], batch size: 17, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:45:30,751 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-02-09 02:45:38,411 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233114.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:45:43,344 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.753e+02 2.536e+02 3.042e+02 3.988e+02 8.675e+02, threshold=6.084e+02, percent-clipped=8.0 +2023-02-09 02:45:43,365 INFO [train.py:901] (1/4) Epoch 29, batch 6800, loss[loss=0.2006, simple_loss=0.2821, pruned_loss=0.05953, over 8036.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2816, pruned_loss=0.05707, over 1618543.29 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:45:56,010 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233139.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:46:18,744 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9831, 2.1065, 1.8776, 2.8041, 1.2466, 1.6967, 2.0195, 2.1690], + device='cuda:1'), covar=tensor([0.0786, 0.0889, 0.0872, 0.0343, 0.1122, 0.1349, 0.0843, 0.0847], + device='cuda:1'), in_proj_covar=tensor([0.0233, 0.0196, 0.0246, 0.0215, 0.0203, 0.0248, 0.0250, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 02:46:19,947 INFO [train.py:901] (1/4) Epoch 29, batch 6850, loss[loss=0.2109, simple_loss=0.2919, pruned_loss=0.06494, over 8298.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.282, pruned_loss=0.0571, over 1619924.08 frames. ], batch size: 49, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:46:22,009 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-02-09 02:46:34,900 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2071, 0.9843, 1.7164, 1.0036, 1.6162, 1.8617, 1.9569, 1.6180], + device='cuda:1'), covar=tensor([0.0814, 0.1213, 0.0483, 0.1627, 0.1230, 0.0322, 0.0714, 0.0482], + device='cuda:1'), in_proj_covar=tensor([0.0311, 0.0328, 0.0297, 0.0327, 0.0328, 0.0280, 0.0449, 0.0311], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 02:46:45,171 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7516, 2.3997, 1.8196, 2.1084, 1.9939, 1.7960, 1.9841, 2.1385], + device='cuda:1'), covar=tensor([0.1114, 0.0367, 0.1020, 0.0587, 0.0695, 0.1204, 0.0815, 0.0797], + device='cuda:1'), in_proj_covar=tensor([0.0358, 0.0246, 0.0346, 0.0315, 0.0302, 0.0350, 0.0351, 0.0321], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 02:46:46,498 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233209.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:46:48,575 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3139, 2.1186, 2.6516, 2.2511, 2.8275, 2.4278, 2.2250, 1.6703], + device='cuda:1'), covar=tensor([0.6076, 0.5326, 0.2215, 0.3976, 0.2487, 0.3391, 0.2011, 0.5669], + device='cuda:1'), in_proj_covar=tensor([0.0975, 0.1037, 0.0844, 0.1010, 0.1030, 0.0943, 0.0780, 0.0859], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:46:55,110 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.759e+02 2.470e+02 3.068e+02 3.817e+02 7.038e+02, threshold=6.136e+02, percent-clipped=5.0 +2023-02-09 02:46:55,130 INFO [train.py:901] (1/4) Epoch 29, batch 6900, loss[loss=0.2028, simple_loss=0.2823, pruned_loss=0.06166, over 8472.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2829, pruned_loss=0.05756, over 1616916.09 frames. ], batch size: 25, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:47:00,399 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3088, 2.1718, 2.7090, 2.2736, 2.7794, 2.4767, 2.2489, 1.6585], + device='cuda:1'), covar=tensor([0.5920, 0.5346, 0.2340, 0.4471, 0.2694, 0.3379, 0.2150, 0.6005], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1037, 0.0844, 0.1009, 0.1030, 0.0943, 0.0779, 0.0858], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:47:04,467 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233234.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:47:16,275 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233251.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:47:31,602 INFO [train.py:901] (1/4) Epoch 29, batch 6950, loss[loss=0.1911, simple_loss=0.2831, pruned_loss=0.04961, over 8568.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2815, pruned_loss=0.0568, over 1615657.46 frames. ], batch size: 31, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:47:33,664 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-02-09 02:47:35,285 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233276.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:47:35,913 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233277.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:48:07,351 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.388e+02 2.860e+02 3.725e+02 6.106e+02, threshold=5.720e+02, percent-clipped=0.0 +2023-02-09 02:48:07,372 INFO [train.py:901] (1/4) Epoch 29, batch 7000, loss[loss=0.1765, simple_loss=0.2646, pruned_loss=0.04419, over 8133.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2809, pruned_loss=0.05611, over 1614988.44 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:48:30,364 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=233352.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:48:44,307 INFO [train.py:901] (1/4) Epoch 29, batch 7050, loss[loss=0.1946, simple_loss=0.2861, pruned_loss=0.0516, over 8601.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.281, pruned_loss=0.05642, over 1614360.70 frames. ], batch size: 49, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:48:51,037 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=233380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:49:21,596 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7459, 2.1680, 3.2097, 1.6462, 2.3571, 2.1361, 1.8785, 2.4197], + device='cuda:1'), covar=tensor([0.1959, 0.2647, 0.0916, 0.4612, 0.2101, 0.3340, 0.2389, 0.2365], + device='cuda:1'), in_proj_covar=tensor([0.0543, 0.0641, 0.0566, 0.0676, 0.0664, 0.0615, 0.0568, 0.0648], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:49:22,015 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.696e+02 2.400e+02 3.088e+02 3.796e+02 6.683e+02, threshold=6.176e+02, percent-clipped=2.0 +2023-02-09 02:49:22,035 INFO [train.py:901] (1/4) Epoch 29, batch 7100, loss[loss=0.1924, simple_loss=0.2654, pruned_loss=0.05968, over 7799.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2816, pruned_loss=0.05654, over 1615128.16 frames. ], batch size: 19, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:49:54,755 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=233467.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:49:57,410 INFO [train.py:901] (1/4) Epoch 29, batch 7150, loss[loss=0.1838, simple_loss=0.264, pruned_loss=0.05182, over 8343.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2813, pruned_loss=0.05656, over 1615190.33 frames. ], batch size: 26, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:50:09,462 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6847, 1.8138, 1.5934, 2.0907, 1.1661, 1.4391, 1.6985, 1.8358], + device='cuda:1'), covar=tensor([0.0676, 0.0698, 0.0847, 0.0529, 0.0937, 0.1170, 0.0647, 0.0673], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0195, 0.0245, 0.0214, 0.0202, 0.0246, 0.0249, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 02:50:14,761 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=233495.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:50:23,099 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.2107, 4.2012, 3.8223, 1.9349, 3.7108, 3.8307, 3.7338, 3.6313], + device='cuda:1'), covar=tensor([0.0821, 0.0572, 0.1028, 0.4607, 0.1023, 0.1037, 0.1262, 0.0868], + device='cuda:1'), in_proj_covar=tensor([0.0549, 0.0462, 0.0454, 0.0563, 0.0446, 0.0470, 0.0445, 0.0414], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:50:23,177 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7465, 1.4069, 1.7419, 1.3001, 1.0299, 1.4625, 1.6864, 1.4023], + device='cuda:1'), covar=tensor([0.0612, 0.1300, 0.1652, 0.1542, 0.0595, 0.1493, 0.0703, 0.0702], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0155, 0.0191, 0.0163, 0.0102, 0.0165, 0.0114, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:1') +2023-02-09 02:50:34,185 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.538e+02 2.385e+02 2.900e+02 3.377e+02 5.605e+02, threshold=5.800e+02, percent-clipped=0.0 +2023-02-09 02:50:34,205 INFO [train.py:901] (1/4) Epoch 29, batch 7200, loss[loss=0.2193, simple_loss=0.3109, pruned_loss=0.06385, over 8631.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2801, pruned_loss=0.05586, over 1613997.09 frames. ], batch size: 31, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:51:10,563 INFO [train.py:901] (1/4) Epoch 29, batch 7250, loss[loss=0.1825, simple_loss=0.2691, pruned_loss=0.04799, over 7665.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2799, pruned_loss=0.05546, over 1616488.77 frames. ], batch size: 19, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:51:11,748 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.64 vs. limit=2.0 +2023-02-09 02:51:36,848 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6820, 1.5487, 2.3052, 1.3880, 1.2535, 2.2468, 0.4204, 1.3742], + device='cuda:1'), covar=tensor([0.1658, 0.1254, 0.0341, 0.1190, 0.2393, 0.0417, 0.1898, 0.1300], + device='cuda:1'), in_proj_covar=tensor([0.0204, 0.0207, 0.0138, 0.0227, 0.0281, 0.0149, 0.0174, 0.0202], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 02:51:46,281 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.384e+02 2.959e+02 3.398e+02 1.041e+03, threshold=5.918e+02, percent-clipped=4.0 +2023-02-09 02:51:46,302 INFO [train.py:901] (1/4) Epoch 29, batch 7300, loss[loss=0.1895, simple_loss=0.2759, pruned_loss=0.05153, over 8144.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2802, pruned_loss=0.05583, over 1620407.65 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:51:46,382 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=233621.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:52:13,017 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233657.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:52:22,375 INFO [train.py:901] (1/4) Epoch 29, batch 7350, loss[loss=0.1944, simple_loss=0.2807, pruned_loss=0.05408, over 8485.00 frames. ], tot_loss[loss=0.1948, simple_loss=0.279, pruned_loss=0.05527, over 1619015.66 frames. ], batch size: 29, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:52:27,563 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4095, 1.6244, 2.1472, 1.3420, 1.5022, 1.6668, 1.4785, 1.5471], + device='cuda:1'), covar=tensor([0.2173, 0.2915, 0.1131, 0.5200, 0.2335, 0.3830, 0.2701, 0.2370], + device='cuda:1'), in_proj_covar=tensor([0.0546, 0.0644, 0.0569, 0.0678, 0.0667, 0.0619, 0.0570, 0.0651], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:52:28,052 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-02-09 02:52:42,944 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.1033, 1.5688, 4.2711, 1.8378, 3.7968, 3.5713, 3.8903, 3.7707], + device='cuda:1'), covar=tensor([0.0719, 0.4839, 0.0645, 0.4265, 0.1144, 0.1071, 0.0681, 0.0771], + device='cuda:1'), in_proj_covar=tensor([0.0695, 0.0671, 0.0756, 0.0663, 0.0755, 0.0642, 0.0650, 0.0733], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:52:48,258 WARNING [train.py:1067] (1/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-02-09 02:52:58,115 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.683e+02 2.663e+02 3.074e+02 4.100e+02 9.512e+02, threshold=6.147e+02, percent-clipped=7.0 +2023-02-09 02:52:58,135 INFO [train.py:901] (1/4) Epoch 29, batch 7400, loss[loss=0.1636, simple_loss=0.2484, pruned_loss=0.03945, over 7802.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2798, pruned_loss=0.05557, over 1617561.14 frames. ], batch size: 20, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:52:59,684 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233723.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:03,229 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233728.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:08,951 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=233736.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:18,927 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233748.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:21,097 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233751.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:32,421 WARNING [train.py:1067] (1/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-02-09 02:53:36,088 INFO [train.py:901] (1/4) Epoch 29, batch 7450, loss[loss=0.1647, simple_loss=0.2489, pruned_loss=0.04026, over 7697.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2792, pruned_loss=0.05537, over 1618274.61 frames. ], batch size: 18, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:53:39,814 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233776.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:42,689 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.8681, 1.6643, 2.0635, 1.6713, 1.1512, 1.7382, 2.3247, 1.9902], + device='cuda:1'), covar=tensor([0.0471, 0.1260, 0.1664, 0.1481, 0.0600, 0.1454, 0.0615, 0.0640], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0155, 0.0191, 0.0163, 0.0102, 0.0165, 0.0115, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:1') +2023-02-09 02:54:12,671 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.619e+02 2.228e+02 2.637e+02 3.402e+02 7.399e+02, threshold=5.273e+02, percent-clipped=2.0 +2023-02-09 02:54:12,691 INFO [train.py:901] (1/4) Epoch 29, batch 7500, loss[loss=0.2001, simple_loss=0.2766, pruned_loss=0.06184, over 7646.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2802, pruned_loss=0.05578, over 1612878.64 frames. ], batch size: 19, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:54:48,746 INFO [train.py:901] (1/4) Epoch 29, batch 7550, loss[loss=0.2394, simple_loss=0.3248, pruned_loss=0.07704, over 8503.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2807, pruned_loss=0.05631, over 1612350.17 frames. ], batch size: 28, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:55:00,147 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0978, 2.2597, 1.8181, 2.9466, 1.2767, 1.6688, 1.9746, 2.2693], + device='cuda:1'), covar=tensor([0.0681, 0.0735, 0.0907, 0.0320, 0.1214, 0.1259, 0.0908, 0.0749], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0195, 0.0244, 0.0213, 0.0202, 0.0246, 0.0249, 0.0205], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 02:55:24,355 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.668e+02 2.424e+02 2.935e+02 3.534e+02 7.288e+02, threshold=5.870e+02, percent-clipped=3.0 +2023-02-09 02:55:24,375 INFO [train.py:901] (1/4) Epoch 29, batch 7600, loss[loss=0.2141, simple_loss=0.3057, pruned_loss=0.06122, over 8629.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2806, pruned_loss=0.0561, over 1613517.63 frames. ], batch size: 34, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:55:24,644 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7858, 1.6374, 2.4937, 1.9340, 2.1004, 1.7796, 1.6213, 1.0602], + device='cuda:1'), covar=tensor([0.7715, 0.6241, 0.2340, 0.4205, 0.3277, 0.4801, 0.3088, 0.5738], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1037, 0.0845, 0.1009, 0.1033, 0.0945, 0.0781, 0.0859], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 02:56:01,024 INFO [train.py:901] (1/4) Epoch 29, batch 7650, loss[loss=0.179, simple_loss=0.2671, pruned_loss=0.04546, over 8136.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2793, pruned_loss=0.05542, over 1613505.16 frames. ], batch size: 22, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:56:12,957 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.7651, 1.5176, 3.9392, 1.6118, 3.4989, 3.2767, 3.5713, 3.5013], + device='cuda:1'), covar=tensor([0.0781, 0.4638, 0.0768, 0.4381, 0.1182, 0.1138, 0.0736, 0.0773], + device='cuda:1'), in_proj_covar=tensor([0.0693, 0.0670, 0.0754, 0.0662, 0.0751, 0.0641, 0.0649, 0.0728], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 02:56:16,597 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233992.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:56:23,674 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234001.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:56:35,522 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234017.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:56:38,038 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.742e+02 2.208e+02 2.731e+02 3.331e+02 6.993e+02, threshold=5.462e+02, percent-clipped=2.0 +2023-02-09 02:56:38,058 INFO [train.py:901] (1/4) Epoch 29, batch 7700, loss[loss=0.1491, simple_loss=0.239, pruned_loss=0.02962, over 8084.00 frames. ], tot_loss[loss=0.1941, simple_loss=0.2785, pruned_loss=0.05482, over 1616098.02 frames. ], batch size: 21, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:56:49,501 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-02-09 02:56:53,146 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234043.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:57:12,872 INFO [train.py:901] (1/4) Epoch 29, batch 7750, loss[loss=0.1756, simple_loss=0.2658, pruned_loss=0.04276, over 8199.00 frames. ], tot_loss[loss=0.1948, simple_loss=0.2791, pruned_loss=0.05523, over 1618554.54 frames. ], batch size: 23, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:57:13,650 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234072.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:57:45,242 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234116.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:57:48,459 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.737e+02 2.355e+02 2.809e+02 3.505e+02 7.382e+02, threshold=5.617e+02, percent-clipped=2.0 +2023-02-09 02:57:48,479 INFO [train.py:901] (1/4) Epoch 29, batch 7800, loss[loss=0.1683, simple_loss=0.2598, pruned_loss=0.03845, over 7813.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2799, pruned_loss=0.05568, over 1616590.15 frames. ], batch size: 20, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:58:24,431 INFO [train.py:901] (1/4) Epoch 29, batch 7850, loss[loss=0.1745, simple_loss=0.2503, pruned_loss=0.04937, over 7420.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2798, pruned_loss=0.0554, over 1614045.91 frames. ], batch size: 17, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:58:36,023 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234187.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:58:58,930 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234220.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:58:59,498 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.576e+02 2.264e+02 2.745e+02 3.500e+02 8.048e+02, threshold=5.490e+02, percent-clipped=4.0 +2023-02-09 02:58:59,518 INFO [train.py:901] (1/4) Epoch 29, batch 7900, loss[loss=0.1822, simple_loss=0.273, pruned_loss=0.04569, over 8200.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.28, pruned_loss=0.05555, over 1617310.01 frames. ], batch size: 23, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:59:19,147 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234249.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:59:34,076 INFO [train.py:901] (1/4) Epoch 29, batch 7950, loss[loss=0.201, simple_loss=0.2804, pruned_loss=0.06074, over 7538.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2808, pruned_loss=0.0561, over 1614860.23 frames. ], batch size: 18, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:59:40,221 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.06 vs. limit=5.0 +2023-02-09 03:00:10,405 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.560e+02 2.366e+02 2.676e+02 3.633e+02 8.832e+02, threshold=5.352e+02, percent-clipped=6.0 +2023-02-09 03:00:10,425 INFO [train.py:901] (1/4) Epoch 29, batch 8000, loss[loss=0.2021, simple_loss=0.3027, pruned_loss=0.05077, over 8690.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2799, pruned_loss=0.05542, over 1614472.08 frames. ], batch size: 34, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 03:00:44,624 INFO [train.py:901] (1/4) Epoch 29, batch 8050, loss[loss=0.1949, simple_loss=0.256, pruned_loss=0.06687, over 6783.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2791, pruned_loss=0.05622, over 1585591.66 frames. ], batch size: 15, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 03:00:45,524 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234372.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:00:55,735 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234387.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:01:02,574 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234397.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:01:19,720 WARNING [train.py:1067] (1/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-09 03:01:23,762 INFO [train.py:901] (1/4) Epoch 30, batch 0, loss[loss=0.2403, simple_loss=0.3177, pruned_loss=0.08151, over 8632.00 frames. ], tot_loss[loss=0.2403, simple_loss=0.3177, pruned_loss=0.08151, over 8632.00 frames. ], batch size: 39, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:01:23,762 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 03:01:35,938 INFO [train.py:935] (1/4) Epoch 30, validation: loss=0.1704, simple_loss=0.27, pruned_loss=0.03537, over 944034.00 frames. +2023-02-09 03:01:35,940 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 03:01:47,831 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.533e+02 2.332e+02 2.743e+02 3.464e+02 7.498e+02, threshold=5.486e+02, percent-clipped=3.0 +2023-02-09 03:01:52,025 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-02-09 03:02:04,488 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234443.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:02:12,494 INFO [train.py:901] (1/4) Epoch 30, batch 50, loss[loss=0.1816, simple_loss=0.2727, pruned_loss=0.04528, over 8615.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2853, pruned_loss=0.06021, over 363475.76 frames. ], batch size: 34, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:02:23,084 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234468.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:02:28,111 WARNING [train.py:1067] (1/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-02-09 03:02:49,595 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234502.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:02:50,896 INFO [train.py:901] (1/4) Epoch 30, batch 100, loss[loss=0.2183, simple_loss=0.3055, pruned_loss=0.06558, over 8436.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2799, pruned_loss=0.05586, over 638730.58 frames. ], batch size: 27, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:02:54,610 WARNING [train.py:1067] (1/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-02-09 03:03:03,348 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.539e+02 2.294e+02 2.794e+02 3.449e+02 7.855e+02, threshold=5.588e+02, percent-clipped=7.0 +2023-02-09 03:03:28,105 INFO [train.py:901] (1/4) Epoch 30, batch 150, loss[loss=0.2084, simple_loss=0.2897, pruned_loss=0.06357, over 8466.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2799, pruned_loss=0.05559, over 855098.37 frames. ], batch size: 27, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:03:35,360 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234564.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:03:56,911 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234593.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:04:04,594 INFO [train.py:901] (1/4) Epoch 30, batch 200, loss[loss=0.2026, simple_loss=0.2659, pruned_loss=0.06962, over 7679.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2802, pruned_loss=0.05617, over 1018631.40 frames. ], batch size: 18, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:04:13,772 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3840, 2.2420, 2.7550, 2.4035, 2.7920, 2.3990, 2.3253, 1.8752], + device='cuda:1'), covar=tensor([0.5113, 0.4780, 0.2166, 0.3974, 0.2633, 0.3296, 0.1885, 0.5357], + device='cuda:1'), in_proj_covar=tensor([0.0973, 0.1039, 0.0847, 0.1012, 0.1034, 0.0945, 0.0781, 0.0861], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:04:16,945 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.375e+02 2.329e+02 2.797e+02 3.759e+02 1.341e+03, threshold=5.593e+02, percent-clipped=8.0 +2023-02-09 03:04:40,208 INFO [train.py:901] (1/4) Epoch 30, batch 250, loss[loss=0.2019, simple_loss=0.2967, pruned_loss=0.05356, over 8467.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2821, pruned_loss=0.05664, over 1152826.29 frames. ], batch size: 49, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:04:48,445 WARNING [train.py:1067] (1/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-02-09 03:04:50,779 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8793, 1.7619, 2.5063, 1.5498, 1.3686, 2.5033, 0.5607, 1.4806], + device='cuda:1'), covar=tensor([0.1462, 0.1104, 0.0329, 0.1126, 0.2349, 0.0366, 0.1752, 0.1345], + device='cuda:1'), in_proj_covar=tensor([0.0203, 0.0208, 0.0139, 0.0227, 0.0280, 0.0149, 0.0174, 0.0203], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 03:04:57,626 WARNING [train.py:1067] (1/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-02-09 03:04:58,540 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234679.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:04,642 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234687.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:16,612 INFO [train.py:901] (1/4) Epoch 30, batch 300, loss[loss=0.2095, simple_loss=0.2996, pruned_loss=0.05967, over 8469.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2806, pruned_loss=0.05543, over 1255419.62 frames. ], batch size: 25, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:05:19,746 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234708.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:29,385 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.381e+02 2.888e+02 3.640e+02 7.253e+02, threshold=5.776e+02, percent-clipped=4.0 +2023-02-09 03:05:45,724 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234744.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:46,513 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6887, 2.0188, 2.0530, 1.4652, 2.1167, 1.5871, 0.5952, 1.9508], + device='cuda:1'), covar=tensor([0.0635, 0.0345, 0.0345, 0.0583, 0.0430, 0.1021, 0.1035, 0.0307], + device='cuda:1'), in_proj_covar=tensor([0.0477, 0.0416, 0.0372, 0.0465, 0.0400, 0.0555, 0.0405, 0.0443], + device='cuda:1'), out_proj_covar=tensor([1.2618e-04, 1.0768e-04, 9.6869e-05, 1.2151e-04, 1.0482e-04, 1.5420e-04, + 1.0793e-04, 1.1575e-04], device='cuda:1') +2023-02-09 03:05:52,652 INFO [train.py:901] (1/4) Epoch 30, batch 350, loss[loss=0.1691, simple_loss=0.2424, pruned_loss=0.04796, over 7700.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2826, pruned_loss=0.05685, over 1334439.22 frames. ], batch size: 18, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:05:55,585 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234758.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:06:14,218 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234783.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:06:27,360 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234800.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:06:30,081 INFO [train.py:901] (1/4) Epoch 30, batch 400, loss[loss=0.2154, simple_loss=0.3002, pruned_loss=0.06531, over 8345.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2842, pruned_loss=0.05797, over 1397492.24 frames. ], batch size: 25, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:06:42,234 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.545e+02 2.513e+02 2.926e+02 3.697e+02 1.204e+03, threshold=5.852e+02, percent-clipped=7.0 +2023-02-09 03:07:06,512 INFO [train.py:901] (1/4) Epoch 30, batch 450, loss[loss=0.2136, simple_loss=0.3005, pruned_loss=0.06329, over 8034.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.283, pruned_loss=0.05737, over 1446058.74 frames. ], batch size: 22, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:07:38,965 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234900.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:07:42,320 INFO [train.py:901] (1/4) Epoch 30, batch 500, loss[loss=0.1888, simple_loss=0.282, pruned_loss=0.04777, over 8291.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2812, pruned_loss=0.05639, over 1484775.32 frames. ], batch size: 23, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:07:54,764 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.531e+02 2.393e+02 2.948e+02 3.833e+02 6.284e+02, threshold=5.896e+02, percent-clipped=1.0 +2023-02-09 03:08:04,859 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234935.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:18,293 INFO [train.py:901] (1/4) Epoch 30, batch 550, loss[loss=0.1707, simple_loss=0.2566, pruned_loss=0.04236, over 7652.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2805, pruned_loss=0.05627, over 1510495.19 frames. ], batch size: 19, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:08:22,852 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234960.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:26,367 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234964.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:34,787 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234976.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:36,243 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234978.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:44,116 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234989.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:54,539 INFO [train.py:901] (1/4) Epoch 30, batch 600, loss[loss=0.1669, simple_loss=0.2614, pruned_loss=0.03619, over 8099.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2802, pruned_loss=0.05578, over 1533013.76 frames. ], batch size: 21, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:09:06,004 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.582e+02 2.485e+02 2.961e+02 3.544e+02 6.861e+02, threshold=5.922e+02, percent-clipped=1.0 +2023-02-09 03:09:10,828 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-02-09 03:09:13,652 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235031.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:09:29,246 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.86 vs. limit=5.0 +2023-02-09 03:09:29,966 INFO [train.py:901] (1/4) Epoch 30, batch 650, loss[loss=0.1964, simple_loss=0.2834, pruned_loss=0.05472, over 8142.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2806, pruned_loss=0.05601, over 1552392.52 frames. ], batch size: 22, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:09:54,554 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235088.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:10:05,672 INFO [train.py:901] (1/4) Epoch 30, batch 700, loss[loss=0.2366, simple_loss=0.3224, pruned_loss=0.07538, over 8809.00 frames. ], tot_loss[loss=0.197, simple_loss=0.281, pruned_loss=0.05648, over 1563447.09 frames. ], batch size: 49, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:10:17,689 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.543e+02 2.477e+02 3.055e+02 3.959e+02 7.285e+02, threshold=6.109e+02, percent-clipped=6.0 +2023-02-09 03:10:33,427 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235144.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:10:34,917 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235146.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:10:41,061 INFO [train.py:901] (1/4) Epoch 30, batch 750, loss[loss=0.1793, simple_loss=0.2652, pruned_loss=0.04668, over 8090.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2809, pruned_loss=0.05636, over 1578541.78 frames. ], batch size: 21, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:10:59,027 WARNING [train.py:1067] (1/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-02-09 03:10:59,126 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.2884, 3.1866, 3.0118, 1.5755, 2.9365, 3.0243, 2.8355, 2.8990], + device='cuda:1'), covar=tensor([0.1085, 0.0767, 0.1232, 0.4222, 0.1108, 0.1088, 0.1595, 0.0999], + device='cuda:1'), in_proj_covar=tensor([0.0551, 0.0461, 0.0451, 0.0562, 0.0445, 0.0470, 0.0446, 0.0414], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:11:08,051 WARNING [train.py:1067] (1/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-02-09 03:11:17,084 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235203.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:11:17,649 INFO [train.py:901] (1/4) Epoch 30, batch 800, loss[loss=0.2101, simple_loss=0.2862, pruned_loss=0.06697, over 8034.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2802, pruned_loss=0.05573, over 1586679.17 frames. ], batch size: 22, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:11:30,473 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.721e+02 2.417e+02 2.836e+02 3.328e+02 8.160e+02, threshold=5.671e+02, percent-clipped=2.0 +2023-02-09 03:11:46,749 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235244.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:11:53,486 INFO [train.py:901] (1/4) Epoch 30, batch 850, loss[loss=0.181, simple_loss=0.2526, pruned_loss=0.05465, over 7218.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2805, pruned_loss=0.05608, over 1591198.77 frames. ], batch size: 16, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:11:57,054 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235259.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:12:07,383 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235273.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:12:29,574 INFO [train.py:901] (1/4) Epoch 30, batch 900, loss[loss=0.1821, simple_loss=0.2702, pruned_loss=0.04705, over 8021.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2801, pruned_loss=0.05588, over 1597086.79 frames. ], batch size: 22, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:12:38,254 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-02-09 03:12:41,044 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235320.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:12:41,602 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.438e+02 3.006e+02 3.865e+02 6.238e+02, threshold=6.012e+02, percent-clipped=6.0 +2023-02-09 03:12:42,346 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235322.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:13:02,599 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.1578, 1.6297, 4.3463, 1.6001, 3.8023, 3.6321, 3.8832, 3.8467], + device='cuda:1'), covar=tensor([0.0655, 0.4631, 0.0567, 0.4768, 0.1261, 0.1046, 0.0705, 0.0749], + device='cuda:1'), in_proj_covar=tensor([0.0696, 0.0673, 0.0758, 0.0667, 0.0755, 0.0645, 0.0652, 0.0731], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:13:05,256 INFO [train.py:901] (1/4) Epoch 30, batch 950, loss[loss=0.2201, simple_loss=0.3023, pruned_loss=0.06888, over 8251.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2804, pruned_loss=0.05587, over 1602285.29 frames. ], batch size: 24, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:13:08,759 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235359.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:13:19,186 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9994, 1.7726, 2.0110, 1.8878, 1.9301, 2.0388, 1.9328, 0.8221], + device='cuda:1'), covar=tensor([0.5611, 0.4636, 0.2389, 0.3567, 0.2520, 0.3382, 0.1991, 0.5126], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1040, 0.0849, 0.1011, 0.1033, 0.0946, 0.0779, 0.0862], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:13:21,830 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235378.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:13:32,984 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-02-09 03:13:39,423 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235402.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:13:40,618 INFO [train.py:901] (1/4) Epoch 30, batch 1000, loss[loss=0.1963, simple_loss=0.2839, pruned_loss=0.05428, over 8503.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2801, pruned_loss=0.05561, over 1608308.73 frames. ], batch size: 26, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:13:52,264 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.479e+02 3.055e+02 4.205e+02 7.814e+02, threshold=6.110e+02, percent-clipped=3.0 +2023-02-09 03:13:56,457 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235427.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:02,519 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235435.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:03,943 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235437.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:08,437 WARNING [train.py:1067] (1/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-02-09 03:14:14,981 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-02-09 03:14:15,939 INFO [train.py:901] (1/4) Epoch 30, batch 1050, loss[loss=0.2113, simple_loss=0.2848, pruned_loss=0.06888, over 7926.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2816, pruned_loss=0.05672, over 1615227.77 frames. ], batch size: 20, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:14:16,045 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.8399, 3.8216, 3.4613, 1.7332, 3.4025, 3.5198, 3.3925, 3.3891], + device='cuda:1'), covar=tensor([0.0952, 0.0636, 0.1138, 0.4971, 0.0988, 0.0970, 0.1406, 0.0886], + device='cuda:1'), in_proj_covar=tensor([0.0550, 0.0459, 0.0450, 0.0562, 0.0444, 0.0469, 0.0444, 0.0412], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:14:19,402 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235459.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:21,156 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-02-09 03:14:36,653 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235484.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:50,591 INFO [train.py:901] (1/4) Epoch 30, batch 1100, loss[loss=0.1522, simple_loss=0.2341, pruned_loss=0.03516, over 7243.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2814, pruned_loss=0.05624, over 1615858.18 frames. ], batch size: 16, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:14:59,086 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235515.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:15:03,870 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.639e+02 2.444e+02 3.135e+02 3.900e+02 6.752e+02, threshold=6.270e+02, percent-clipped=2.0 +2023-02-09 03:15:11,142 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8347, 2.0298, 2.1213, 1.4164, 2.2656, 1.5181, 0.7881, 2.0208], + device='cuda:1'), covar=tensor([0.0736, 0.0422, 0.0342, 0.0745, 0.0504, 0.1048, 0.1146, 0.0395], + device='cuda:1'), in_proj_covar=tensor([0.0483, 0.0422, 0.0375, 0.0471, 0.0404, 0.0561, 0.0410, 0.0448], + device='cuda:1'), out_proj_covar=tensor([1.2799e-04, 1.0900e-04, 9.7585e-05, 1.2291e-04, 1.0577e-04, 1.5615e-04, + 1.0930e-04, 1.1730e-04], device='cuda:1') +2023-02-09 03:15:17,338 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235540.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:15:26,821 INFO [train.py:901] (1/4) Epoch 30, batch 1150, loss[loss=0.2288, simple_loss=0.3031, pruned_loss=0.07719, over 8559.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2808, pruned_loss=0.05605, over 1615611.61 frames. ], batch size: 31, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:15:35,848 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-02-09 03:16:03,362 INFO [train.py:901] (1/4) Epoch 30, batch 1200, loss[loss=0.2317, simple_loss=0.3005, pruned_loss=0.0815, over 7071.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2798, pruned_loss=0.05554, over 1611479.74 frames. ], batch size: 72, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:16:11,496 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235615.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:16:12,764 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235617.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:16:16,062 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.630e+02 2.378e+02 2.945e+02 3.640e+02 8.540e+02, threshold=5.890e+02, percent-clipped=4.0 +2023-02-09 03:16:30,035 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235640.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:16:39,500 INFO [train.py:901] (1/4) Epoch 30, batch 1250, loss[loss=0.1697, simple_loss=0.2457, pruned_loss=0.04681, over 7798.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.28, pruned_loss=0.05585, over 1610450.54 frames. ], batch size: 19, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:16:43,892 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.41 vs. limit=5.0 +2023-02-09 03:17:06,188 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235691.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:07,625 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235693.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:10,371 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235697.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:15,249 INFO [train.py:901] (1/4) Epoch 30, batch 1300, loss[loss=0.2129, simple_loss=0.2923, pruned_loss=0.0668, over 8559.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2799, pruned_loss=0.05595, over 1612438.41 frames. ], batch size: 31, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:17:23,826 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235716.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:25,224 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235718.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:27,370 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8874, 1.6291, 2.0042, 1.8248, 1.9829, 1.9368, 1.7952, 0.9337], + device='cuda:1'), covar=tensor([0.6176, 0.5030, 0.2169, 0.3366, 0.2544, 0.3175, 0.1997, 0.5157], + device='cuda:1'), in_proj_covar=tensor([0.0964, 0.1030, 0.0841, 0.1002, 0.1025, 0.0939, 0.0773, 0.0855], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:17:27,734 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.629e+02 2.444e+02 2.774e+02 3.314e+02 6.214e+02, threshold=5.548e+02, percent-clipped=2.0 +2023-02-09 03:17:27,830 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235722.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:17:34,611 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235732.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:35,435 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.91 vs. limit=5.0 +2023-02-09 03:17:50,118 INFO [train.py:901] (1/4) Epoch 30, batch 1350, loss[loss=0.2301, simple_loss=0.3164, pruned_loss=0.07189, over 8078.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2803, pruned_loss=0.05636, over 1611856.12 frames. ], batch size: 21, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:18:26,587 INFO [train.py:901] (1/4) Epoch 30, batch 1400, loss[loss=0.2005, simple_loss=0.2888, pruned_loss=0.05615, over 8464.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2804, pruned_loss=0.05615, over 1615237.79 frames. ], batch size: 25, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:18:30,246 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235809.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:18:39,097 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.499e+02 2.338e+02 2.741e+02 3.583e+02 7.907e+02, threshold=5.482e+02, percent-clipped=6.0 +2023-02-09 03:18:49,551 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235837.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:19:00,850 INFO [train.py:901] (1/4) Epoch 30, batch 1450, loss[loss=0.1882, simple_loss=0.2721, pruned_loss=0.05217, over 8322.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2812, pruned_loss=0.05655, over 1612215.26 frames. ], batch size: 25, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:19:07,596 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-02-09 03:19:22,541 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.66 vs. limit=2.0 +2023-02-09 03:19:38,200 INFO [train.py:901] (1/4) Epoch 30, batch 1500, loss[loss=0.1992, simple_loss=0.2861, pruned_loss=0.05619, over 8600.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2803, pruned_loss=0.05621, over 1612906.34 frames. ], batch size: 31, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:19:42,078 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-02-09 03:19:51,220 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.723e+02 2.306e+02 2.900e+02 3.560e+02 8.272e+02, threshold=5.801e+02, percent-clipped=7.0 +2023-02-09 03:19:56,726 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-02-09 03:20:14,191 INFO [train.py:901] (1/4) Epoch 30, batch 1550, loss[loss=0.1654, simple_loss=0.2501, pruned_loss=0.04032, over 7931.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2795, pruned_loss=0.05524, over 1616972.12 frames. ], batch size: 20, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:20:17,111 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9759, 1.3081, 4.3793, 1.8590, 2.4596, 4.9438, 5.1100, 4.3298], + device='cuda:1'), covar=tensor([0.1272, 0.2164, 0.0253, 0.1930, 0.1185, 0.0187, 0.0350, 0.0535], + device='cuda:1'), in_proj_covar=tensor([0.0309, 0.0329, 0.0296, 0.0329, 0.0327, 0.0282, 0.0447, 0.0310], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 03:20:22,056 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-02-09 03:20:27,550 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8345, 1.6783, 2.3292, 1.5375, 1.4515, 2.3214, 0.8273, 1.6094], + device='cuda:1'), covar=tensor([0.1319, 0.1116, 0.0333, 0.1019, 0.2105, 0.0372, 0.1625, 0.1322], + device='cuda:1'), in_proj_covar=tensor([0.0203, 0.0208, 0.0139, 0.0226, 0.0280, 0.0149, 0.0175, 0.0202], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 03:20:38,870 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235988.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:20:51,290 INFO [train.py:901] (1/4) Epoch 30, batch 1600, loss[loss=0.213, simple_loss=0.2928, pruned_loss=0.06662, over 8707.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2798, pruned_loss=0.05602, over 1612836.85 frames. ], batch size: 34, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:20:57,833 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236013.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:21:00,082 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6685, 2.5886, 1.8893, 2.3892, 2.2872, 1.7172, 2.1876, 2.2460], + device='cuda:1'), covar=tensor([0.1622, 0.0436, 0.1319, 0.0730, 0.0787, 0.1604, 0.1130, 0.1207], + device='cuda:1'), in_proj_covar=tensor([0.0360, 0.0246, 0.0349, 0.0317, 0.0304, 0.0352, 0.0354, 0.0325], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 03:21:04,932 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.669e+02 2.693e+02 3.134e+02 4.092e+02 8.333e+02, threshold=6.267e+02, percent-clipped=7.0 +2023-02-09 03:21:15,453 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=236036.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:21:19,088 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=236041.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:21:28,375 INFO [train.py:901] (1/4) Epoch 30, batch 1650, loss[loss=0.2464, simple_loss=0.3238, pruned_loss=0.08455, over 8793.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2809, pruned_loss=0.05648, over 1615341.60 frames. ], batch size: 30, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:21:56,788 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236093.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:22:04,819 INFO [train.py:901] (1/4) Epoch 30, batch 1700, loss[loss=0.19, simple_loss=0.2738, pruned_loss=0.05311, over 8659.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2799, pruned_loss=0.05602, over 1614246.06 frames. ], batch size: 34, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:22:15,382 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236118.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:22:17,735 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.438e+02 2.804e+02 3.459e+02 5.840e+02, threshold=5.608e+02, percent-clipped=0.0 +2023-02-09 03:22:34,511 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.57 vs. limit=5.0 +2023-02-09 03:22:40,514 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=236153.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:22:41,159 INFO [train.py:901] (1/4) Epoch 30, batch 1750, loss[loss=0.208, simple_loss=0.2938, pruned_loss=0.06111, over 8282.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2794, pruned_loss=0.05585, over 1612033.75 frames. ], batch size: 23, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:22:42,647 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=236156.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:23:07,920 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.1658, 1.8849, 2.2689, 2.0297, 2.2960, 2.2425, 2.1033, 1.2357], + device='cuda:1'), covar=tensor([0.6077, 0.5246, 0.2346, 0.4049, 0.2686, 0.3353, 0.1944, 0.5681], + device='cuda:1'), in_proj_covar=tensor([0.0972, 0.1037, 0.0848, 0.1008, 0.1032, 0.0946, 0.0778, 0.0860], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:23:16,214 INFO [train.py:901] (1/4) Epoch 30, batch 1800, loss[loss=0.2166, simple_loss=0.2998, pruned_loss=0.06671, over 8326.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2805, pruned_loss=0.05637, over 1614181.84 frames. ], batch size: 26, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:23:19,551 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-02-09 03:23:29,387 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.538e+02 2.503e+02 3.165e+02 3.852e+02 7.294e+02, threshold=6.329e+02, percent-clipped=5.0 +2023-02-09 03:23:31,000 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0119, 1.8651, 2.2916, 2.0561, 2.2698, 2.1026, 1.9442, 1.1661], + device='cuda:1'), covar=tensor([0.5786, 0.4762, 0.2083, 0.3596, 0.2524, 0.3044, 0.1960, 0.5089], + device='cuda:1'), in_proj_covar=tensor([0.0970, 0.1035, 0.0846, 0.1005, 0.1030, 0.0945, 0.0776, 0.0858], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:23:52,507 INFO [train.py:901] (1/4) Epoch 30, batch 1850, loss[loss=0.2698, simple_loss=0.3361, pruned_loss=0.1017, over 8037.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2814, pruned_loss=0.05701, over 1609464.61 frames. ], batch size: 22, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:24:03,962 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=236268.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:24:28,815 INFO [train.py:901] (1/4) Epoch 30, batch 1900, loss[loss=0.2639, simple_loss=0.3319, pruned_loss=0.09792, over 7012.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2803, pruned_loss=0.05589, over 1609195.02 frames. ], batch size: 71, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:24:41,167 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.66 vs. limit=2.0 +2023-02-09 03:24:41,427 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.517e+02 2.325e+02 3.004e+02 3.832e+02 8.674e+02, threshold=6.008e+02, percent-clipped=3.0 +2023-02-09 03:25:01,724 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-02-09 03:25:05,198 INFO [train.py:901] (1/4) Epoch 30, batch 1950, loss[loss=0.2028, simple_loss=0.2906, pruned_loss=0.0575, over 8626.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2796, pruned_loss=0.0558, over 1607616.54 frames. ], batch size: 34, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:25:13,614 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-02-09 03:25:24,199 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=236380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:25:25,016 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.4642, 2.6992, 2.9344, 1.8536, 3.0928, 2.0342, 1.5577, 2.4628], + device='cuda:1'), covar=tensor([0.0948, 0.0431, 0.0388, 0.0853, 0.0594, 0.0944, 0.1126, 0.0565], + device='cuda:1'), in_proj_covar=tensor([0.0477, 0.0418, 0.0373, 0.0465, 0.0400, 0.0555, 0.0405, 0.0442], + device='cuda:1'), out_proj_covar=tensor([1.2627e-04, 1.0814e-04, 9.7134e-05, 1.2130e-04, 1.0478e-04, 1.5435e-04, + 1.0794e-04, 1.1561e-04], device='cuda:1') +2023-02-09 03:25:25,037 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9132, 2.4360, 4.0263, 1.7841, 2.9242, 2.5063, 2.0197, 2.9639], + device='cuda:1'), covar=tensor([0.1856, 0.2643, 0.0811, 0.4669, 0.2041, 0.3176, 0.2386, 0.2454], + device='cuda:1'), in_proj_covar=tensor([0.0546, 0.0646, 0.0569, 0.0681, 0.0670, 0.0622, 0.0573, 0.0651], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:25:33,317 WARNING [train.py:1067] (1/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-02-09 03:25:41,055 INFO [train.py:901] (1/4) Epoch 30, batch 2000, loss[loss=0.2228, simple_loss=0.3101, pruned_loss=0.06776, over 8428.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2814, pruned_loss=0.05669, over 1612982.30 frames. ], batch size: 49, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:25:43,357 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.5495, 2.3322, 2.9697, 2.4442, 3.0063, 2.5585, 2.4850, 2.0552], + device='cuda:1'), covar=tensor([0.5905, 0.5540, 0.2410, 0.4527, 0.3071, 0.3500, 0.1916, 0.5863], + device='cuda:1'), in_proj_covar=tensor([0.0970, 0.1036, 0.0846, 0.1007, 0.1032, 0.0947, 0.0778, 0.0857], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:25:46,881 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236412.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:25:53,568 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.624e+02 2.406e+02 2.956e+02 3.756e+02 9.982e+02, threshold=5.913e+02, percent-clipped=8.0 +2023-02-09 03:26:04,555 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236437.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:26:16,258 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6411, 1.8380, 1.5513, 2.3107, 1.0815, 1.4348, 1.7368, 1.8640], + device='cuda:1'), covar=tensor([0.0777, 0.0709, 0.0969, 0.0388, 0.1024, 0.1208, 0.0685, 0.0707], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0194, 0.0244, 0.0215, 0.0201, 0.0247, 0.0251, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 03:26:16,752 INFO [train.py:901] (1/4) Epoch 30, batch 2050, loss[loss=0.2255, simple_loss=0.306, pruned_loss=0.0725, over 8033.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.281, pruned_loss=0.05658, over 1610228.77 frames. ], batch size: 22, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:26:28,372 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([0.9628, 1.6045, 1.3744, 1.4800, 1.3030, 1.2765, 1.2209, 1.2280], + device='cuda:1'), covar=tensor([0.1326, 0.0576, 0.1483, 0.0661, 0.0862, 0.1678, 0.1151, 0.1018], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0245, 0.0348, 0.0316, 0.0304, 0.0351, 0.0354, 0.0323], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 03:26:47,592 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=236495.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:26:53,532 INFO [train.py:901] (1/4) Epoch 30, batch 2100, loss[loss=0.203, simple_loss=0.2931, pruned_loss=0.05648, over 8501.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2807, pruned_loss=0.05623, over 1612352.01 frames. ], batch size: 26, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:27:06,242 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.478e+02 2.477e+02 3.089e+02 3.892e+02 8.089e+02, threshold=6.178e+02, percent-clipped=3.0 +2023-02-09 03:27:07,829 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236524.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:27:16,738 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3959, 1.2257, 2.3856, 1.3472, 2.2797, 2.5541, 2.7361, 2.1584], + device='cuda:1'), covar=tensor([0.1177, 0.1554, 0.0402, 0.2098, 0.0718, 0.0382, 0.0584, 0.0659], + device='cuda:1'), in_proj_covar=tensor([0.0310, 0.0330, 0.0298, 0.0329, 0.0329, 0.0282, 0.0449, 0.0312], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 03:27:25,243 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236549.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:27:28,497 INFO [train.py:901] (1/4) Epoch 30, batch 2150, loss[loss=0.1915, simple_loss=0.2713, pruned_loss=0.05589, over 8125.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2793, pruned_loss=0.05588, over 1606341.41 frames. ], batch size: 22, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:27:59,354 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=3.84 vs. limit=5.0 +2023-02-09 03:28:04,587 INFO [train.py:901] (1/4) Epoch 30, batch 2200, loss[loss=0.2174, simple_loss=0.303, pruned_loss=0.06585, over 8630.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2791, pruned_loss=0.05595, over 1603490.27 frames. ], batch size: 31, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:28:06,910 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.7066, 4.7671, 4.3072, 2.0619, 4.2077, 4.3884, 4.2324, 4.2384], + device='cuda:1'), covar=tensor([0.0703, 0.0474, 0.0971, 0.4941, 0.0912, 0.0851, 0.1253, 0.0626], + device='cuda:1'), in_proj_covar=tensor([0.0552, 0.0462, 0.0450, 0.0565, 0.0447, 0.0474, 0.0450, 0.0415], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:28:18,768 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.663e+02 2.435e+02 2.816e+02 3.564e+02 9.413e+02, threshold=5.632e+02, percent-clipped=3.0 +2023-02-09 03:28:40,981 INFO [train.py:901] (1/4) Epoch 30, batch 2250, loss[loss=0.1784, simple_loss=0.2594, pruned_loss=0.04873, over 8076.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2796, pruned_loss=0.05582, over 1605946.31 frames. ], batch size: 21, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:29:04,195 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-02-09 03:29:16,926 INFO [train.py:901] (1/4) Epoch 30, batch 2300, loss[loss=0.2113, simple_loss=0.3033, pruned_loss=0.05968, over 8510.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2807, pruned_loss=0.05613, over 1611641.64 frames. ], batch size: 26, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:29:29,254 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.819e+02 2.489e+02 3.036e+02 4.215e+02 7.962e+02, threshold=6.071e+02, percent-clipped=6.0 +2023-02-09 03:29:50,614 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236751.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:29:52,574 INFO [train.py:901] (1/4) Epoch 30, batch 2350, loss[loss=0.1962, simple_loss=0.2703, pruned_loss=0.06106, over 7537.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2801, pruned_loss=0.05575, over 1611804.78 frames. ], batch size: 18, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:30:08,614 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236776.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:30:29,051 INFO [train.py:901] (1/4) Epoch 30, batch 2400, loss[loss=0.1842, simple_loss=0.2602, pruned_loss=0.05409, over 8091.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2798, pruned_loss=0.05563, over 1613849.59 frames. ], batch size: 21, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:30:42,217 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.757e+02 2.281e+02 2.685e+02 3.729e+02 8.099e+02, threshold=5.371e+02, percent-clipped=9.0 +2023-02-09 03:31:05,192 INFO [train.py:901] (1/4) Epoch 30, batch 2450, loss[loss=0.1642, simple_loss=0.2466, pruned_loss=0.04088, over 7823.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2793, pruned_loss=0.05584, over 1618223.26 frames. ], batch size: 20, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:31:39,710 INFO [train.py:901] (1/4) Epoch 30, batch 2500, loss[loss=0.2336, simple_loss=0.3268, pruned_loss=0.07022, over 8326.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2792, pruned_loss=0.05556, over 1623317.47 frames. ], batch size: 25, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:31:43,942 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=236910.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:31:52,882 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.501e+02 2.373e+02 3.086e+02 3.767e+02 7.222e+02, threshold=6.171e+02, percent-clipped=6.0 +2023-02-09 03:32:13,801 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2241, 3.6912, 2.4005, 2.9577, 2.8906, 2.0935, 2.8889, 3.0697], + device='cuda:1'), covar=tensor([0.1719, 0.0342, 0.1183, 0.0847, 0.0878, 0.1533, 0.1182, 0.1123], + device='cuda:1'), in_proj_covar=tensor([0.0356, 0.0242, 0.0345, 0.0314, 0.0302, 0.0350, 0.0351, 0.0320], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 03:32:16,579 INFO [train.py:901] (1/4) Epoch 30, batch 2550, loss[loss=0.2058, simple_loss=0.2963, pruned_loss=0.05761, over 8524.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2791, pruned_loss=0.05557, over 1621857.70 frames. ], batch size: 28, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:32:54,184 INFO [train.py:901] (1/4) Epoch 30, batch 2600, loss[loss=0.1726, simple_loss=0.2479, pruned_loss=0.04865, over 7542.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2794, pruned_loss=0.05596, over 1617818.69 frames. ], batch size: 18, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:33:06,912 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.551e+02 2.458e+02 3.021e+02 3.974e+02 8.394e+02, threshold=6.042e+02, percent-clipped=5.0 +2023-02-09 03:33:12,329 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=4.20 vs. limit=5.0 +2023-02-09 03:33:30,295 INFO [train.py:901] (1/4) Epoch 30, batch 2650, loss[loss=0.2024, simple_loss=0.2805, pruned_loss=0.0622, over 7809.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2801, pruned_loss=0.05654, over 1617198.44 frames. ], batch size: 20, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:34:06,433 INFO [train.py:901] (1/4) Epoch 30, batch 2700, loss[loss=0.2104, simple_loss=0.2841, pruned_loss=0.0683, over 7802.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2796, pruned_loss=0.05574, over 1617721.87 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:34:07,585 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-02-09 03:34:18,974 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.680e+02 2.464e+02 3.015e+02 4.068e+02 7.247e+02, threshold=6.030e+02, percent-clipped=1.0 +2023-02-09 03:34:41,475 INFO [train.py:901] (1/4) Epoch 30, batch 2750, loss[loss=0.2001, simple_loss=0.2876, pruned_loss=0.05629, over 8454.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2792, pruned_loss=0.05587, over 1611989.69 frames. ], batch size: 25, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:35:18,232 INFO [train.py:901] (1/4) Epoch 30, batch 2800, loss[loss=0.2099, simple_loss=0.2959, pruned_loss=0.06197, over 8500.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2798, pruned_loss=0.05593, over 1615713.01 frames. ], batch size: 26, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:35:20,498 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.6165, 1.5643, 1.9457, 1.5748, 1.1128, 1.6162, 2.2187, 1.9313], + device='cuda:1'), covar=tensor([0.0491, 0.1242, 0.1581, 0.1430, 0.0580, 0.1489, 0.0619, 0.0642], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0155, 0.0192, 0.0163, 0.0102, 0.0165, 0.0114, 0.0149], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:1') +2023-02-09 03:35:31,339 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.593e+02 2.300e+02 2.824e+02 3.573e+02 8.919e+02, threshold=5.648e+02, percent-clipped=3.0 +2023-02-09 03:35:45,297 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.4908, 1.7256, 4.6785, 1.9196, 4.1804, 3.9278, 4.2386, 4.1565], + device='cuda:1'), covar=tensor([0.0521, 0.4184, 0.0519, 0.4209, 0.0936, 0.0866, 0.0545, 0.0560], + device='cuda:1'), in_proj_covar=tensor([0.0694, 0.0671, 0.0752, 0.0670, 0.0756, 0.0645, 0.0655, 0.0729], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:35:48,278 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-09 03:35:53,018 INFO [train.py:901] (1/4) Epoch 30, batch 2850, loss[loss=0.2034, simple_loss=0.2895, pruned_loss=0.05862, over 8530.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2808, pruned_loss=0.05609, over 1620034.24 frames. ], batch size: 28, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:35:53,088 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=237254.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:36:18,912 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-02-09 03:36:27,596 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.66 vs. limit=2.0 +2023-02-09 03:36:29,225 INFO [train.py:901] (1/4) Epoch 30, batch 2900, loss[loss=0.2085, simple_loss=0.2916, pruned_loss=0.0627, over 8187.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2814, pruned_loss=0.05646, over 1616145.89 frames. ], batch size: 23, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:36:42,591 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.704e+02 2.592e+02 3.021e+02 4.387e+02 8.419e+02, threshold=6.042e+02, percent-clipped=5.0 +2023-02-09 03:37:04,019 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-02-09 03:37:05,361 INFO [train.py:901] (1/4) Epoch 30, batch 2950, loss[loss=0.209, simple_loss=0.2974, pruned_loss=0.06028, over 8134.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2809, pruned_loss=0.05598, over 1613564.72 frames. ], batch size: 22, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:37:15,788 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=237369.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:37:36,194 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.4164, 4.3709, 4.0072, 1.7580, 3.9007, 4.0637, 3.9700, 3.8776], + device='cuda:1'), covar=tensor([0.0686, 0.0500, 0.0874, 0.4888, 0.0862, 0.0882, 0.1181, 0.0782], + device='cuda:1'), in_proj_covar=tensor([0.0550, 0.0461, 0.0450, 0.0562, 0.0445, 0.0473, 0.0447, 0.0415], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:37:40,336 INFO [train.py:901] (1/4) Epoch 30, batch 3000, loss[loss=0.2384, simple_loss=0.3142, pruned_loss=0.0813, over 7150.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2819, pruned_loss=0.05614, over 1617308.45 frames. ], batch size: 72, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:37:40,336 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 03:37:54,066 INFO [train.py:935] (1/4) Epoch 30, validation: loss=0.1704, simple_loss=0.2697, pruned_loss=0.0356, over 944034.00 frames. +2023-02-09 03:37:54,068 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 03:38:07,361 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.697e+02 2.369e+02 2.918e+02 3.560e+02 6.316e+02, threshold=5.836e+02, percent-clipped=1.0 +2023-02-09 03:38:31,180 INFO [train.py:901] (1/4) Epoch 30, batch 3050, loss[loss=0.2209, simple_loss=0.3, pruned_loss=0.07089, over 8511.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2826, pruned_loss=0.05683, over 1617184.89 frames. ], batch size: 26, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:39:01,958 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.1002, 1.2307, 1.2146, 0.9091, 1.1907, 1.0299, 0.1550, 1.1988], + device='cuda:1'), covar=tensor([0.0474, 0.0453, 0.0433, 0.0580, 0.0514, 0.1025, 0.1020, 0.0375], + device='cuda:1'), in_proj_covar=tensor([0.0483, 0.0422, 0.0376, 0.0467, 0.0403, 0.0562, 0.0409, 0.0446], + device='cuda:1'), out_proj_covar=tensor([1.2778e-04, 1.0912e-04, 9.7950e-05, 1.2182e-04, 1.0531e-04, 1.5642e-04, + 1.0891e-04, 1.1676e-04], device='cuda:1') +2023-02-09 03:39:07,118 INFO [train.py:901] (1/4) Epoch 30, batch 3100, loss[loss=0.2204, simple_loss=0.3058, pruned_loss=0.06749, over 8462.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.284, pruned_loss=0.05784, over 1620642.22 frames. ], batch size: 27, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:39:10,846 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237509.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:39:14,042 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.35 vs. limit=2.0 +2023-02-09 03:39:19,646 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.408e+02 2.429e+02 3.016e+02 3.485e+02 6.483e+02, threshold=6.032e+02, percent-clipped=4.0 +2023-02-09 03:39:20,786 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-02-09 03:39:43,982 INFO [train.py:901] (1/4) Epoch 30, batch 3150, loss[loss=0.1394, simple_loss=0.225, pruned_loss=0.0269, over 7420.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.282, pruned_loss=0.05674, over 1621157.06 frames. ], batch size: 17, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:40:03,668 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237581.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:40:21,025 INFO [train.py:901] (1/4) Epoch 30, batch 3200, loss[loss=0.1716, simple_loss=0.259, pruned_loss=0.04206, over 7968.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2799, pruned_loss=0.05587, over 1618623.15 frames. ], batch size: 21, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:40:33,293 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.611e+02 2.320e+02 2.861e+02 3.592e+02 8.186e+02, threshold=5.722e+02, percent-clipped=5.0 +2023-02-09 03:40:35,497 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=237625.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:40:52,664 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=237650.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:40:56,035 INFO [train.py:901] (1/4) Epoch 30, batch 3250, loss[loss=0.1809, simple_loss=0.2721, pruned_loss=0.04489, over 8331.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2808, pruned_loss=0.05649, over 1617168.69 frames. ], batch size: 25, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:41:32,173 INFO [train.py:901] (1/4) Epoch 30, batch 3300, loss[loss=0.224, simple_loss=0.3031, pruned_loss=0.07248, over 8027.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2813, pruned_loss=0.05679, over 1616825.86 frames. ], batch size: 22, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:41:44,070 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.40 vs. limit=5.0 +2023-02-09 03:41:45,786 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.739e+02 2.392e+02 2.907e+02 3.818e+02 6.093e+02, threshold=5.813e+02, percent-clipped=2.0 +2023-02-09 03:42:07,968 INFO [train.py:901] (1/4) Epoch 30, batch 3350, loss[loss=0.2616, simple_loss=0.335, pruned_loss=0.09406, over 8439.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2831, pruned_loss=0.05824, over 1613341.44 frames. ], batch size: 29, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:42:44,184 INFO [train.py:901] (1/4) Epoch 30, batch 3400, loss[loss=0.1586, simple_loss=0.26, pruned_loss=0.02856, over 8098.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2818, pruned_loss=0.05773, over 1607978.91 frames. ], batch size: 23, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:42:47,126 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237808.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:42:57,410 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.758e+02 2.484e+02 3.245e+02 4.483e+02 9.283e+02, threshold=6.490e+02, percent-clipped=12.0 +2023-02-09 03:43:00,482 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3675, 2.6897, 2.8445, 1.6912, 3.1676, 1.9905, 1.5556, 2.4707], + device='cuda:1'), covar=tensor([0.1024, 0.0473, 0.0368, 0.1001, 0.0640, 0.1001, 0.1151, 0.0614], + device='cuda:1'), in_proj_covar=tensor([0.0486, 0.0425, 0.0380, 0.0470, 0.0406, 0.0566, 0.0410, 0.0449], + device='cuda:1'), out_proj_covar=tensor([1.2850e-04, 1.0992e-04, 9.8809e-05, 1.2263e-04, 1.0614e-04, 1.5748e-04, + 1.0928e-04, 1.1749e-04], device='cuda:1') +2023-02-09 03:43:19,446 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=237853.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:43:20,122 INFO [train.py:901] (1/4) Epoch 30, batch 3450, loss[loss=0.1908, simple_loss=0.2823, pruned_loss=0.04966, over 8603.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2832, pruned_loss=0.05781, over 1614054.38 frames. ], batch size: 34, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:43:56,089 INFO [train.py:901] (1/4) Epoch 30, batch 3500, loss[loss=0.1978, simple_loss=0.2972, pruned_loss=0.04924, over 8556.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2819, pruned_loss=0.05651, over 1613923.71 frames. ], batch size: 31, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:44:08,739 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.712e+02 2.505e+02 3.010e+02 3.725e+02 8.965e+02, threshold=6.019e+02, percent-clipped=4.0 +2023-02-09 03:44:11,005 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=237925.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:44:11,725 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6720, 1.3476, 2.8927, 1.4714, 2.4408, 3.0822, 3.2882, 2.6489], + device='cuda:1'), covar=tensor([0.1216, 0.1784, 0.0338, 0.2045, 0.0758, 0.0332, 0.0606, 0.0560], + device='cuda:1'), in_proj_covar=tensor([0.0308, 0.0327, 0.0296, 0.0326, 0.0328, 0.0281, 0.0447, 0.0308], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 03:44:14,958 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237930.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:44:22,365 WARNING [train.py:1067] (1/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-02-09 03:44:32,802 INFO [train.py:901] (1/4) Epoch 30, batch 3550, loss[loss=0.1629, simple_loss=0.2493, pruned_loss=0.03831, over 7649.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2817, pruned_loss=0.05649, over 1610454.33 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:44:43,164 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=237968.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:44:49,855 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-09 03:45:09,425 INFO [train.py:901] (1/4) Epoch 30, batch 3600, loss[loss=0.1965, simple_loss=0.282, pruned_loss=0.05545, over 8095.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2816, pruned_loss=0.05644, over 1613150.87 frames. ], batch size: 21, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:45:22,458 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.684e+02 2.286e+02 2.832e+02 3.360e+02 7.556e+02, threshold=5.664e+02, percent-clipped=2.0 +2023-02-09 03:45:35,405 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238040.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:45:40,890 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.5868, 4.5811, 4.1491, 1.9989, 4.0586, 4.2165, 4.0431, 4.0435], + device='cuda:1'), covar=tensor([0.0646, 0.0494, 0.0892, 0.4661, 0.0804, 0.1036, 0.1325, 0.0784], + device='cuda:1'), in_proj_covar=tensor([0.0551, 0.0463, 0.0452, 0.0563, 0.0445, 0.0475, 0.0451, 0.0416], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:45:44,902 INFO [train.py:901] (1/4) Epoch 30, batch 3650, loss[loss=0.1706, simple_loss=0.2671, pruned_loss=0.03707, over 8322.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2792, pruned_loss=0.05541, over 1606961.09 frames. ], batch size: 25, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:46:05,849 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238082.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:12,778 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238092.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:20,739 INFO [train.py:901] (1/4) Epoch 30, batch 3700, loss[loss=0.2164, simple_loss=0.3042, pruned_loss=0.06429, over 8284.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2796, pruned_loss=0.0553, over 1610438.63 frames. ], batch size: 23, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:46:29,085 WARNING [train.py:1067] (1/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-02-09 03:46:33,265 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.392e+02 2.352e+02 3.001e+02 3.686e+02 7.575e+02, threshold=6.003e+02, percent-clipped=3.0 +2023-02-09 03:46:50,316 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238144.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:54,655 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6847, 1.3491, 2.9028, 1.5354, 2.4384, 3.1466, 3.2587, 2.7065], + device='cuda:1'), covar=tensor([0.1222, 0.1809, 0.0320, 0.1978, 0.0749, 0.0290, 0.0672, 0.0570], + device='cuda:1'), in_proj_covar=tensor([0.0310, 0.0330, 0.0298, 0.0328, 0.0331, 0.0283, 0.0451, 0.0310], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 03:46:56,061 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238152.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:57,365 INFO [train.py:901] (1/4) Epoch 30, batch 3750, loss[loss=0.2065, simple_loss=0.2825, pruned_loss=0.06526, over 8462.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2801, pruned_loss=0.05573, over 1610552.65 frames. ], batch size: 25, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:47:19,946 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6128, 2.0890, 3.0547, 1.4521, 2.4032, 2.0392, 1.7148, 2.3986], + device='cuda:1'), covar=tensor([0.2126, 0.2556, 0.1036, 0.5031, 0.1992, 0.3497, 0.2581, 0.2411], + device='cuda:1'), in_proj_covar=tensor([0.0547, 0.0644, 0.0569, 0.0679, 0.0671, 0.0620, 0.0573, 0.0649], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:47:33,683 INFO [train.py:901] (1/4) Epoch 30, batch 3800, loss[loss=0.2096, simple_loss=0.3011, pruned_loss=0.05908, over 8481.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2806, pruned_loss=0.05615, over 1609981.47 frames. ], batch size: 26, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:47:46,040 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.710e+02 2.427e+02 2.911e+02 3.474e+02 7.215e+02, threshold=5.821e+02, percent-clipped=2.0 +2023-02-09 03:47:47,585 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238224.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:47:55,413 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2692, 2.0714, 2.6733, 2.2678, 2.6972, 2.3682, 2.1535, 1.5413], + device='cuda:1'), covar=tensor([0.6018, 0.5115, 0.2136, 0.4215, 0.2642, 0.3379, 0.2141, 0.5819], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1041, 0.0851, 0.1016, 0.1037, 0.0951, 0.0782, 0.0863], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:48:05,483 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238249.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:48:05,502 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.0495, 1.2521, 1.2025, 0.7285, 1.2192, 1.0608, 0.0614, 1.1949], + device='cuda:1'), covar=tensor([0.0516, 0.0421, 0.0405, 0.0678, 0.0489, 0.1037, 0.1007, 0.0373], + device='cuda:1'), in_proj_covar=tensor([0.0482, 0.0422, 0.0378, 0.0468, 0.0405, 0.0561, 0.0408, 0.0446], + device='cuda:1'), out_proj_covar=tensor([1.2749e-04, 1.0912e-04, 9.8345e-05, 1.2207e-04, 1.0605e-04, 1.5607e-04, + 1.0864e-04, 1.1678e-04], device='cuda:1') +2023-02-09 03:48:09,471 INFO [train.py:901] (1/4) Epoch 30, batch 3850, loss[loss=0.1854, simple_loss=0.2605, pruned_loss=0.05517, over 8241.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2817, pruned_loss=0.05672, over 1613784.92 frames. ], batch size: 22, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:48:18,888 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238267.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:48:23,698 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238274.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:48:37,707 WARNING [train.py:1067] (1/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-02-09 03:48:39,907 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238296.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:48:45,176 INFO [train.py:901] (1/4) Epoch 30, batch 3900, loss[loss=0.2054, simple_loss=0.2776, pruned_loss=0.06656, over 6009.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2795, pruned_loss=0.05596, over 1609245.17 frames. ], batch size: 13, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:48:57,800 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238321.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:48:58,294 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.463e+02 2.361e+02 2.887e+02 3.538e+02 6.169e+02, threshold=5.773e+02, percent-clipped=2.0 +2023-02-09 03:49:20,478 INFO [train.py:901] (1/4) Epoch 30, batch 3950, loss[loss=0.1847, simple_loss=0.26, pruned_loss=0.05467, over 7280.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2805, pruned_loss=0.05614, over 1609298.22 frames. ], batch size: 16, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:49:46,426 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238389.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:49:56,957 INFO [train.py:901] (1/4) Epoch 30, batch 4000, loss[loss=0.1692, simple_loss=0.2558, pruned_loss=0.04128, over 8083.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2802, pruned_loss=0.05587, over 1610859.68 frames. ], batch size: 21, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:50:09,960 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.496e+02 2.354e+02 2.920e+02 3.674e+02 8.815e+02, threshold=5.839e+02, percent-clipped=5.0 +2023-02-09 03:50:12,702 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238426.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:50:20,360 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238436.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:50:21,771 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238438.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:50:32,677 INFO [train.py:901] (1/4) Epoch 30, batch 4050, loss[loss=0.1716, simple_loss=0.2543, pruned_loss=0.04446, over 7642.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2808, pruned_loss=0.05625, over 1609655.99 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:50:57,404 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238488.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:09,340 INFO [train.py:901] (1/4) Epoch 30, batch 4100, loss[loss=0.2408, simple_loss=0.3139, pruned_loss=0.08383, over 7242.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2811, pruned_loss=0.05665, over 1606990.11 frames. ], batch size: 73, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:51:21,790 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.512e+02 2.287e+02 2.925e+02 3.934e+02 1.031e+03, threshold=5.850e+02, percent-clipped=7.0 +2023-02-09 03:51:22,755 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238523.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:35,957 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238541.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:41,373 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238548.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:43,263 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238551.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:45,004 INFO [train.py:901] (1/4) Epoch 30, batch 4150, loss[loss=0.1925, simple_loss=0.2678, pruned_loss=0.05863, over 6829.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2814, pruned_loss=0.05678, over 1609262.73 frames. ], batch size: 15, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:52:19,900 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238603.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:52:20,442 INFO [train.py:901] (1/4) Epoch 30, batch 4200, loss[loss=0.1816, simple_loss=0.2497, pruned_loss=0.05676, over 7244.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2818, pruned_loss=0.05742, over 1603646.38 frames. ], batch size: 16, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:52:33,705 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.657e+02 2.494e+02 3.256e+02 4.447e+02 1.288e+03, threshold=6.511e+02, percent-clipped=8.0 +2023-02-09 03:52:41,497 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-02-09 03:52:50,658 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238645.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:52:56,580 INFO [train.py:901] (1/4) Epoch 30, batch 4250, loss[loss=0.2019, simple_loss=0.2834, pruned_loss=0.06021, over 8497.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2812, pruned_loss=0.05675, over 1608011.88 frames. ], batch size: 29, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:53:02,319 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.13 vs. limit=2.0 +2023-02-09 03:53:05,078 WARNING [train.py:1067] (1/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-02-09 03:53:07,961 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238670.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:53:25,820 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7088, 1.4702, 1.7653, 1.3857, 0.9311, 1.4908, 1.5622, 1.4024], + device='cuda:1'), covar=tensor([0.0598, 0.1276, 0.1607, 0.1408, 0.0622, 0.1458, 0.0737, 0.0679], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0154, 0.0191, 0.0162, 0.0102, 0.0164, 0.0113, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 03:53:27,923 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238699.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:53:29,417 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5116, 1.8601, 2.6793, 1.4410, 1.9951, 1.8378, 1.6209, 2.0204], + device='cuda:1'), covar=tensor([0.1904, 0.2760, 0.0913, 0.4812, 0.1968, 0.3582, 0.2503, 0.2390], + device='cuda:1'), in_proj_covar=tensor([0.0546, 0.0644, 0.0569, 0.0679, 0.0671, 0.0619, 0.0572, 0.0650], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:53:31,133 INFO [train.py:901] (1/4) Epoch 30, batch 4300, loss[loss=0.2264, simple_loss=0.3013, pruned_loss=0.07572, over 8139.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2815, pruned_loss=0.05668, over 1608015.87 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:53:44,805 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.604e+02 2.303e+02 2.743e+02 3.342e+02 6.438e+02, threshold=5.486e+02, percent-clipped=0.0 +2023-02-09 03:54:06,887 INFO [train.py:901] (1/4) Epoch 30, batch 4350, loss[loss=0.225, simple_loss=0.3084, pruned_loss=0.0708, over 8592.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2823, pruned_loss=0.05711, over 1614456.11 frames. ], batch size: 31, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:54:27,360 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238782.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:54:36,398 WARNING [train.py:1067] (1/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-02-09 03:54:37,978 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238797.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:54:42,562 INFO [train.py:901] (1/4) Epoch 30, batch 4400, loss[loss=0.1973, simple_loss=0.2882, pruned_loss=0.05322, over 8703.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2813, pruned_loss=0.05655, over 1613913.17 frames. ], batch size: 50, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:54:44,935 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238807.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:54:55,879 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.538e+02 2.573e+02 3.023e+02 3.983e+02 6.680e+02, threshold=6.046e+02, percent-clipped=2.0 +2023-02-09 03:54:56,101 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238822.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:03,795 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238832.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:15,276 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-02-09 03:55:18,601 INFO [train.py:901] (1/4) Epoch 30, batch 4450, loss[loss=0.237, simple_loss=0.3097, pruned_loss=0.08209, over 8357.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2812, pruned_loss=0.057, over 1612877.47 frames. ], batch size: 24, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:55:22,523 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238859.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:26,395 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-02-09 03:55:40,870 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238884.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:50,464 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238897.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:55:55,009 INFO [train.py:901] (1/4) Epoch 30, batch 4500, loss[loss=0.2221, simple_loss=0.3005, pruned_loss=0.07183, over 7818.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2816, pruned_loss=0.05736, over 1611604.29 frames. ], batch size: 20, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:56:07,440 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.647e+02 2.335e+02 2.828e+02 3.474e+02 8.376e+02, threshold=5.656e+02, percent-clipped=3.0 +2023-02-09 03:56:08,192 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-02-09 03:56:11,163 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0690, 2.1847, 1.8533, 2.9498, 1.3361, 1.7110, 2.1239, 2.1270], + device='cuda:1'), covar=tensor([0.0709, 0.0816, 0.0823, 0.0324, 0.1131, 0.1302, 0.0828, 0.0895], + device='cuda:1'), in_proj_covar=tensor([0.0232, 0.0195, 0.0245, 0.0215, 0.0203, 0.0247, 0.0250, 0.0206], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 03:56:30,982 INFO [train.py:901] (1/4) Epoch 30, batch 4550, loss[loss=0.1857, simple_loss=0.2737, pruned_loss=0.04878, over 8043.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2812, pruned_loss=0.05672, over 1612779.39 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:56:31,132 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238954.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:56:37,965 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-02-09 03:56:38,545 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7990, 1.4414, 1.7780, 1.4188, 0.9343, 1.5145, 1.6238, 1.6493], + device='cuda:1'), covar=tensor([0.0589, 0.1332, 0.1610, 0.1477, 0.0611, 0.1459, 0.0730, 0.0623], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0155, 0.0192, 0.0163, 0.0102, 0.0165, 0.0114, 0.0149], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:1') +2023-02-09 03:57:06,050 INFO [train.py:901] (1/4) Epoch 30, batch 4600, loss[loss=0.2224, simple_loss=0.3036, pruned_loss=0.07059, over 8464.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2809, pruned_loss=0.05679, over 1612786.20 frames. ], batch size: 27, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:57:19,157 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.347e+02 2.832e+02 3.443e+02 5.144e+02, threshold=5.665e+02, percent-clipped=0.0 +2023-02-09 03:57:20,227 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.75 vs. limit=5.0 +2023-02-09 03:57:34,218 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=239043.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:57:41,731 INFO [train.py:901] (1/4) Epoch 30, batch 4650, loss[loss=0.1905, simple_loss=0.2705, pruned_loss=0.05523, over 7550.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2825, pruned_loss=0.05725, over 1616391.32 frames. ], batch size: 18, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:57:46,069 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3915, 2.2031, 2.9908, 2.3507, 2.7729, 2.4569, 2.2709, 1.6448], + device='cuda:1'), covar=tensor([0.5989, 0.5514, 0.2101, 0.4274, 0.3001, 0.3379, 0.2112, 0.6099], + device='cuda:1'), in_proj_covar=tensor([0.0972, 0.1038, 0.0851, 0.1013, 0.1035, 0.0949, 0.0783, 0.0861], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 03:57:58,377 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([3.8126, 1.4945, 4.0087, 1.5125, 3.5616, 3.3732, 3.6474, 3.5481], + device='cuda:1'), covar=tensor([0.0772, 0.4488, 0.0705, 0.4353, 0.1238, 0.1035, 0.0652, 0.0806], + device='cuda:1'), in_proj_covar=tensor([0.0700, 0.0674, 0.0759, 0.0672, 0.0758, 0.0649, 0.0658, 0.0733], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 03:58:03,513 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0245, 1.5481, 1.7916, 1.4785, 0.9841, 1.5821, 1.8632, 1.7113], + device='cuda:1'), covar=tensor([0.0562, 0.1279, 0.1719, 0.1470, 0.0603, 0.1491, 0.0682, 0.0650], + device='cuda:1'), in_proj_covar=tensor([0.0100, 0.0155, 0.0192, 0.0163, 0.0102, 0.0165, 0.0114, 0.0149], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:1') +2023-02-09 03:58:17,765 INFO [train.py:901] (1/4) Epoch 30, batch 4700, loss[loss=0.2679, simple_loss=0.3469, pruned_loss=0.09441, over 8783.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2823, pruned_loss=0.05741, over 1612326.72 frames. ], batch size: 30, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:58:30,951 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.697e+02 2.353e+02 2.866e+02 3.941e+02 8.957e+02, threshold=5.733e+02, percent-clipped=8.0 +2023-02-09 03:58:52,011 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-02-09 03:58:52,478 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=239153.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:58:52,925 INFO [train.py:901] (1/4) Epoch 30, batch 4750, loss[loss=0.2204, simple_loss=0.3127, pruned_loss=0.06398, over 8589.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.283, pruned_loss=0.05777, over 1612117.68 frames. ], batch size: 31, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:58:55,954 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=239158.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:58:58,518 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8054, 1.3590, 3.3473, 1.5854, 2.3317, 3.6512, 3.7509, 3.1430], + device='cuda:1'), covar=tensor([0.1274, 0.1992, 0.0305, 0.1958, 0.1038, 0.0242, 0.0546, 0.0488], + device='cuda:1'), in_proj_covar=tensor([0.0309, 0.0329, 0.0296, 0.0329, 0.0330, 0.0282, 0.0450, 0.0309], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 03:59:05,427 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3565, 3.6514, 2.6844, 3.2422, 3.0178, 2.1633, 3.1167, 3.2306], + device='cuda:1'), covar=tensor([0.1693, 0.0480, 0.1085, 0.0705, 0.0855, 0.1583, 0.1063, 0.1120], + device='cuda:1'), in_proj_covar=tensor([0.0360, 0.0246, 0.0347, 0.0317, 0.0304, 0.0351, 0.0351, 0.0323], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 03:59:10,882 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=239178.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:59:12,054 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-02-09 03:59:14,171 WARNING [train.py:1067] (1/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-02-09 03:59:28,642 INFO [train.py:901] (1/4) Epoch 30, batch 4800, loss[loss=0.2162, simple_loss=0.2837, pruned_loss=0.07437, over 7796.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2807, pruned_loss=0.05708, over 1608916.73 frames. ], batch size: 20, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:59:41,698 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.637e+02 2.393e+02 3.010e+02 3.751e+02 7.640e+02, threshold=6.020e+02, percent-clipped=2.0 +2023-02-09 04:00:04,581 INFO [train.py:901] (1/4) Epoch 30, batch 4850, loss[loss=0.1723, simple_loss=0.2529, pruned_loss=0.04588, over 8090.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2798, pruned_loss=0.05625, over 1606998.71 frames. ], batch size: 21, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:00:06,723 WARNING [train.py:1067] (1/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-02-09 04:00:36,498 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=239298.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:00:40,436 INFO [train.py:901] (1/4) Epoch 30, batch 4900, loss[loss=0.2328, simple_loss=0.317, pruned_loss=0.07427, over 8367.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2799, pruned_loss=0.05633, over 1608380.38 frames. ], batch size: 24, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:00:53,055 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.481e+02 2.412e+02 2.818e+02 3.519e+02 1.028e+03, threshold=5.635e+02, percent-clipped=4.0 +2023-02-09 04:01:04,924 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.5903, 4.5591, 4.1846, 2.0673, 4.0596, 4.1810, 4.1123, 4.1042], + device='cuda:1'), covar=tensor([0.0646, 0.0447, 0.0846, 0.4379, 0.0876, 0.0951, 0.1149, 0.0690], + device='cuda:1'), in_proj_covar=tensor([0.0556, 0.0465, 0.0454, 0.0568, 0.0450, 0.0479, 0.0453, 0.0418], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:01:15,933 INFO [train.py:901] (1/4) Epoch 30, batch 4950, loss[loss=0.2027, simple_loss=0.2932, pruned_loss=0.05613, over 8097.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2814, pruned_loss=0.05707, over 1609126.45 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:01:51,704 INFO [train.py:901] (1/4) Epoch 30, batch 5000, loss[loss=0.286, simple_loss=0.3484, pruned_loss=0.1118, over 8619.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2811, pruned_loss=0.05673, over 1612773.79 frames. ], batch size: 34, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:01:58,097 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=239413.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:01:58,889 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=239414.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:02:04,994 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.597e+02 2.513e+02 3.095e+02 3.810e+02 1.179e+03, threshold=6.190e+02, percent-clipped=9.0 +2023-02-09 04:02:07,460 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6968, 2.6006, 1.9468, 2.4453, 2.2459, 1.7441, 2.2068, 2.2409], + device='cuda:1'), covar=tensor([0.1557, 0.0478, 0.1229, 0.0607, 0.0767, 0.1453, 0.1039, 0.1136], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0246, 0.0345, 0.0315, 0.0302, 0.0348, 0.0349, 0.0322], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 04:02:17,689 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=239439.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:02:19,893 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([0.9928, 1.5780, 1.3673, 1.5247, 1.3087, 1.3073, 1.2992, 1.2690], + device='cuda:1'), covar=tensor([0.1282, 0.0568, 0.1534, 0.0630, 0.0822, 0.1645, 0.1005, 0.0928], + device='cuda:1'), in_proj_covar=tensor([0.0359, 0.0246, 0.0345, 0.0315, 0.0302, 0.0348, 0.0349, 0.0321], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 04:02:29,032 INFO [train.py:901] (1/4) Epoch 30, batch 5050, loss[loss=0.191, simple_loss=0.2717, pruned_loss=0.05519, over 8132.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2808, pruned_loss=0.05637, over 1612198.67 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:02:52,709 WARNING [train.py:1067] (1/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-02-09 04:03:05,838 INFO [train.py:901] (1/4) Epoch 30, batch 5100, loss[loss=0.2165, simple_loss=0.3071, pruned_loss=0.06296, over 8289.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2815, pruned_loss=0.05666, over 1615121.31 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:03:20,035 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.557e+02 2.525e+02 3.230e+02 3.994e+02 1.175e+03, threshold=6.461e+02, percent-clipped=6.0 +2023-02-09 04:03:42,221 INFO [train.py:901] (1/4) Epoch 30, batch 5150, loss[loss=0.2026, simple_loss=0.2828, pruned_loss=0.06121, over 8009.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2821, pruned_loss=0.05733, over 1616016.40 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:03:53,438 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.8533, 1.8797, 2.1881, 1.9089, 1.1947, 1.8852, 2.5035, 2.2813], + device='cuda:1'), covar=tensor([0.0510, 0.1126, 0.1564, 0.1328, 0.0553, 0.1309, 0.0557, 0.0574], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0155, 0.0192, 0.0163, 0.0102, 0.0165, 0.0114, 0.0149], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:1') +2023-02-09 04:04:18,743 INFO [train.py:901] (1/4) Epoch 30, batch 5200, loss[loss=0.1866, simple_loss=0.2806, pruned_loss=0.04631, over 8099.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2813, pruned_loss=0.05658, over 1614908.57 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:04:31,919 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.498e+02 2.360e+02 2.794e+02 3.430e+02 1.458e+03, threshold=5.587e+02, percent-clipped=2.0 +2023-02-09 04:04:44,695 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=239640.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:04:54,717 INFO [train.py:901] (1/4) Epoch 30, batch 5250, loss[loss=0.1615, simple_loss=0.2378, pruned_loss=0.04267, over 7531.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2797, pruned_loss=0.05605, over 1610552.29 frames. ], batch size: 18, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:04:56,121 WARNING [train.py:1067] (1/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-02-09 04:05:05,209 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=239669.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:05:23,273 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=239694.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:05:29,842 INFO [train.py:901] (1/4) Epoch 30, batch 5300, loss[loss=0.1935, simple_loss=0.2897, pruned_loss=0.04865, over 8253.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2797, pruned_loss=0.05589, over 1612435.53 frames. ], batch size: 24, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:05:43,733 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.433e+02 2.937e+02 3.850e+02 7.663e+02, threshold=5.875e+02, percent-clipped=5.0 +2023-02-09 04:05:46,043 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9648, 2.1016, 1.8191, 2.6330, 1.1814, 1.6974, 1.9952, 2.0340], + device='cuda:1'), covar=tensor([0.0713, 0.0697, 0.0808, 0.0355, 0.1015, 0.1177, 0.0732, 0.0783], + device='cuda:1'), in_proj_covar=tensor([0.0230, 0.0194, 0.0243, 0.0213, 0.0201, 0.0245, 0.0248, 0.0203], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 04:05:54,319 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6752, 1.9697, 2.8880, 1.5690, 2.2802, 2.0684, 1.7650, 2.3367], + device='cuda:1'), covar=tensor([0.1921, 0.2744, 0.1055, 0.4809, 0.1995, 0.3321, 0.2577, 0.2310], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0645, 0.0568, 0.0679, 0.0669, 0.0619, 0.0572, 0.0649], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:06:04,873 INFO [train.py:901] (1/4) Epoch 30, batch 5350, loss[loss=0.1639, simple_loss=0.2465, pruned_loss=0.0406, over 8088.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2807, pruned_loss=0.05631, over 1618284.89 frames. ], batch size: 21, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:06:22,341 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.2374, 2.2277, 1.7690, 1.9823, 1.7203, 1.6035, 1.6465, 1.6713], + device='cuda:1'), covar=tensor([0.1343, 0.0478, 0.1280, 0.0547, 0.0819, 0.1531, 0.0990, 0.0878], + device='cuda:1'), in_proj_covar=tensor([0.0358, 0.0245, 0.0345, 0.0315, 0.0301, 0.0348, 0.0349, 0.0320], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 04:06:40,969 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5914, 1.4081, 1.6787, 1.3143, 0.9400, 1.4597, 1.5584, 1.2371], + device='cuda:1'), covar=tensor([0.0605, 0.1279, 0.1740, 0.1508, 0.0585, 0.1485, 0.0697, 0.0736], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0154, 0.0191, 0.0163, 0.0102, 0.0165, 0.0113, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 04:06:41,534 INFO [train.py:901] (1/4) Epoch 30, batch 5400, loss[loss=0.2013, simple_loss=0.2916, pruned_loss=0.05545, over 8294.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2808, pruned_loss=0.05646, over 1615580.93 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:06:55,040 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.816e+02 2.371e+02 2.881e+02 3.522e+02 8.420e+02, threshold=5.763e+02, percent-clipped=7.0 +2023-02-09 04:07:17,634 INFO [train.py:901] (1/4) Epoch 30, batch 5450, loss[loss=0.195, simple_loss=0.2792, pruned_loss=0.0554, over 8516.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2809, pruned_loss=0.05679, over 1616326.49 frames. ], batch size: 49, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:07:49,889 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-02-09 04:07:53,966 INFO [train.py:901] (1/4) Epoch 30, batch 5500, loss[loss=0.2186, simple_loss=0.3012, pruned_loss=0.06801, over 8333.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2798, pruned_loss=0.05599, over 1616828.57 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:08:08,775 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.536e+02 2.378e+02 3.012e+02 4.037e+02 9.246e+02, threshold=6.023e+02, percent-clipped=5.0 +2023-02-09 04:08:30,327 INFO [train.py:901] (1/4) Epoch 30, batch 5550, loss[loss=0.2096, simple_loss=0.2917, pruned_loss=0.06369, over 8655.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2799, pruned_loss=0.05594, over 1618591.72 frames. ], batch size: 39, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:08:50,819 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=239984.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:09:07,157 INFO [train.py:901] (1/4) Epoch 30, batch 5600, loss[loss=0.1941, simple_loss=0.2863, pruned_loss=0.05096, over 7801.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2796, pruned_loss=0.0561, over 1615123.27 frames. ], batch size: 20, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:09:15,197 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-02-09 04:09:21,041 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.553e+02 2.479e+02 2.939e+02 3.472e+02 8.474e+02, threshold=5.878e+02, percent-clipped=2.0 +2023-02-09 04:09:21,976 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7661, 1.6381, 2.1826, 1.3567, 1.3280, 2.1067, 0.3791, 1.3646], + device='cuda:1'), covar=tensor([0.1221, 0.1125, 0.0320, 0.0826, 0.2114, 0.0411, 0.1620, 0.1106], + device='cuda:1'), in_proj_covar=tensor([0.0204, 0.0206, 0.0138, 0.0224, 0.0279, 0.0149, 0.0175, 0.0202], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 04:09:42,585 INFO [train.py:901] (1/4) Epoch 30, batch 5650, loss[loss=0.1891, simple_loss=0.2669, pruned_loss=0.05569, over 8483.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2794, pruned_loss=0.05625, over 1610459.82 frames. ], batch size: 28, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:09:44,731 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.7909, 5.9617, 5.2655, 2.4983, 5.2260, 5.6000, 5.4579, 5.4859], + device='cuda:1'), covar=tensor([0.0515, 0.0303, 0.0768, 0.4337, 0.0745, 0.0857, 0.0887, 0.0591], + device='cuda:1'), in_proj_covar=tensor([0.0558, 0.0466, 0.0456, 0.0572, 0.0453, 0.0480, 0.0456, 0.0419], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:09:59,646 WARNING [train.py:1067] (1/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-02-09 04:10:14,055 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=240099.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:10:17,325 INFO [train.py:901] (1/4) Epoch 30, batch 5700, loss[loss=0.2492, simple_loss=0.3206, pruned_loss=0.08888, over 7025.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2803, pruned_loss=0.05648, over 1611066.84 frames. ], batch size: 72, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:10:20,763 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.4629, 1.7624, 1.4237, 2.8166, 1.1935, 1.3121, 2.1206, 1.8955], + device='cuda:1'), covar=tensor([0.1481, 0.1277, 0.1749, 0.0377, 0.1316, 0.1976, 0.0906, 0.0998], + device='cuda:1'), in_proj_covar=tensor([0.0230, 0.0193, 0.0243, 0.0213, 0.0201, 0.0245, 0.0247, 0.0203], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 04:10:31,996 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.343e+02 2.506e+02 3.162e+02 4.194e+02 1.225e+03, threshold=6.325e+02, percent-clipped=8.0 +2023-02-09 04:10:53,042 INFO [train.py:901] (1/4) Epoch 30, batch 5750, loss[loss=0.2127, simple_loss=0.2989, pruned_loss=0.06329, over 8512.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.28, pruned_loss=0.05676, over 1609822.28 frames. ], batch size: 28, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:11:04,196 WARNING [train.py:1067] (1/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-02-09 04:11:21,800 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.0628, 1.6432, 1.3762, 1.5671, 1.2983, 1.2633, 1.2583, 1.3631], + device='cuda:1'), covar=tensor([0.1221, 0.0563, 0.1541, 0.0636, 0.0929, 0.1724, 0.1127, 0.0889], + device='cuda:1'), in_proj_covar=tensor([0.0357, 0.0245, 0.0344, 0.0314, 0.0301, 0.0346, 0.0349, 0.0318], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 04:11:28,591 INFO [train.py:901] (1/4) Epoch 30, batch 5800, loss[loss=0.1809, simple_loss=0.2655, pruned_loss=0.04819, over 8519.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2806, pruned_loss=0.05712, over 1610945.73 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:11:35,160 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.92 vs. limit=5.0 +2023-02-09 04:11:39,938 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.8090, 2.0015, 1.6733, 2.5408, 1.1572, 1.4901, 1.9198, 1.9927], + device='cuda:1'), covar=tensor([0.0751, 0.0776, 0.0904, 0.0387, 0.1117, 0.1359, 0.0747, 0.0764], + device='cuda:1'), in_proj_covar=tensor([0.0230, 0.0194, 0.0244, 0.0213, 0.0201, 0.0245, 0.0247, 0.0203], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 04:11:42,452 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.461e+02 2.400e+02 2.667e+02 3.487e+02 8.848e+02, threshold=5.334e+02, percent-clipped=2.0 +2023-02-09 04:12:04,252 INFO [train.py:901] (1/4) Epoch 30, batch 5850, loss[loss=0.1798, simple_loss=0.2611, pruned_loss=0.04926, over 7783.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2818, pruned_loss=0.05697, over 1614978.43 frames. ], batch size: 19, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:12:33,881 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5846, 1.4636, 1.8548, 1.2296, 1.3094, 1.7900, 0.2565, 1.2083], + device='cuda:1'), covar=tensor([0.1414, 0.1249, 0.0427, 0.0841, 0.2209, 0.0499, 0.1733, 0.1168], + device='cuda:1'), in_proj_covar=tensor([0.0205, 0.0208, 0.0139, 0.0225, 0.0281, 0.0150, 0.0176, 0.0203], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 04:12:39,826 INFO [train.py:901] (1/4) Epoch 30, batch 5900, loss[loss=0.1936, simple_loss=0.2778, pruned_loss=0.05471, over 8367.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2822, pruned_loss=0.05741, over 1613794.05 frames. ], batch size: 24, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:12:52,667 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=2.75 vs. limit=5.0 +2023-02-09 04:12:53,715 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.373e+02 2.330e+02 2.970e+02 3.920e+02 1.059e+03, threshold=5.939e+02, percent-clipped=6.0 +2023-02-09 04:13:15,506 INFO [train.py:901] (1/4) Epoch 30, batch 5950, loss[loss=0.2099, simple_loss=0.2975, pruned_loss=0.06117, over 8452.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.282, pruned_loss=0.05674, over 1609992.19 frames. ], batch size: 27, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:13:16,423 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=240355.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:13:18,456 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=240358.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:13:33,797 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=240380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:13:50,593 INFO [train.py:901] (1/4) Epoch 30, batch 6000, loss[loss=0.1835, simple_loss=0.2761, pruned_loss=0.04548, over 8332.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2809, pruned_loss=0.05587, over 1606292.35 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:13:50,593 INFO [train.py:926] (1/4) Computing validation loss +2023-02-09 04:14:04,296 INFO [train.py:935] (1/4) Epoch 30, validation: loss=0.1701, simple_loss=0.2695, pruned_loss=0.03536, over 944034.00 frames. +2023-02-09 04:14:04,297 INFO [train.py:936] (1/4) Maximum memory allocated so far is 6668MB +2023-02-09 04:14:17,955 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.732e+02 2.377e+02 3.122e+02 3.554e+02 6.850e+02, threshold=6.243e+02, percent-clipped=2.0 +2023-02-09 04:14:39,925 INFO [train.py:901] (1/4) Epoch 30, batch 6050, loss[loss=0.1995, simple_loss=0.2829, pruned_loss=0.05804, over 8029.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2801, pruned_loss=0.05562, over 1603057.01 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:15:04,207 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-02-09 04:15:16,411 INFO [train.py:901] (1/4) Epoch 30, batch 6100, loss[loss=0.2127, simple_loss=0.2985, pruned_loss=0.06347, over 8320.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2818, pruned_loss=0.05657, over 1606011.76 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:15:30,263 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.619e+02 2.423e+02 2.992e+02 3.767e+02 7.583e+02, threshold=5.983e+02, percent-clipped=4.0 +2023-02-09 04:15:40,564 WARNING [train.py:1067] (1/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-02-09 04:15:51,549 INFO [train.py:901] (1/4) Epoch 30, batch 6150, loss[loss=0.2158, simple_loss=0.3069, pruned_loss=0.06233, over 8456.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2811, pruned_loss=0.0563, over 1609904.64 frames. ], batch size: 27, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:16:28,465 INFO [train.py:901] (1/4) Epoch 30, batch 6200, loss[loss=0.2062, simple_loss=0.2931, pruned_loss=0.0597, over 8643.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2802, pruned_loss=0.05588, over 1611061.70 frames. ], batch size: 34, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:16:44,278 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.619e+02 2.533e+02 2.968e+02 3.901e+02 6.917e+02, threshold=5.935e+02, percent-clipped=4.0 +2023-02-09 04:17:05,856 INFO [train.py:901] (1/4) Epoch 30, batch 6250, loss[loss=0.2362, simple_loss=0.2999, pruned_loss=0.0863, over 6461.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2792, pruned_loss=0.05564, over 1606996.63 frames. ], batch size: 71, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:17:34,725 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.9854, 1.6800, 2.2659, 1.6170, 1.1531, 1.9039, 2.4536, 2.4755], + device='cuda:1'), covar=tensor([0.0461, 0.1115, 0.1403, 0.1372, 0.0532, 0.1329, 0.0564, 0.0533], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0155, 0.0191, 0.0163, 0.0102, 0.0165, 0.0113, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 04:17:39,922 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=240702.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:17:41,234 INFO [train.py:901] (1/4) Epoch 30, batch 6300, loss[loss=0.1883, simple_loss=0.2783, pruned_loss=0.04912, over 8290.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.28, pruned_loss=0.05538, over 1614230.57 frames. ], batch size: 23, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:17:54,925 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.808e+02 2.569e+02 3.101e+02 4.376e+02 1.063e+03, threshold=6.203e+02, percent-clipped=9.0 +2023-02-09 04:18:17,086 INFO [train.py:901] (1/4) Epoch 30, batch 6350, loss[loss=0.1839, simple_loss=0.2701, pruned_loss=0.04887, over 8502.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2812, pruned_loss=0.0557, over 1621343.07 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:18:41,761 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-02-09 04:18:47,105 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3376, 2.5560, 2.8251, 1.9135, 3.1411, 1.8352, 1.6043, 2.3070], + device='cuda:1'), covar=tensor([0.0834, 0.0424, 0.0330, 0.0726, 0.0502, 0.0984, 0.0914, 0.0539], + device='cuda:1'), in_proj_covar=tensor([0.0486, 0.0420, 0.0379, 0.0470, 0.0405, 0.0562, 0.0411, 0.0450], + device='cuda:1'), out_proj_covar=tensor([1.2857e-04, 1.0813e-04, 9.8695e-05, 1.2268e-04, 1.0587e-04, 1.5621e-04, + 1.0949e-04, 1.1748e-04], device='cuda:1') +2023-02-09 04:18:53,320 INFO [train.py:901] (1/4) Epoch 30, batch 6400, loss[loss=0.1811, simple_loss=0.2755, pruned_loss=0.04337, over 8250.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2816, pruned_loss=0.05565, over 1621214.25 frames. ], batch size: 24, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:19:02,602 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=240817.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:19:07,074 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.453e+02 2.422e+02 2.800e+02 3.642e+02 5.918e+02, threshold=5.600e+02, percent-clipped=0.0 +2023-02-09 04:19:28,686 INFO [train.py:901] (1/4) Epoch 30, batch 6450, loss[loss=0.1896, simple_loss=0.279, pruned_loss=0.05007, over 8469.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2817, pruned_loss=0.0559, over 1621213.71 frames. ], batch size: 25, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:19:33,454 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.2294, 2.5311, 2.7280, 1.6603, 2.9686, 1.8017, 1.5037, 2.1789], + device='cuda:1'), covar=tensor([0.0949, 0.0432, 0.0335, 0.0873, 0.0628, 0.0949, 0.1131, 0.0668], + device='cuda:1'), in_proj_covar=tensor([0.0485, 0.0419, 0.0378, 0.0468, 0.0404, 0.0561, 0.0409, 0.0448], + device='cuda:1'), out_proj_covar=tensor([1.2815e-04, 1.0786e-04, 9.8505e-05, 1.2213e-04, 1.0561e-04, 1.5575e-04, + 1.0907e-04, 1.1711e-04], device='cuda:1') +2023-02-09 04:19:38,862 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0332, 1.5215, 3.3337, 1.5522, 2.4240, 3.6072, 3.7858, 3.1161], + device='cuda:1'), covar=tensor([0.1176, 0.1852, 0.0293, 0.1972, 0.0983, 0.0221, 0.0446, 0.0503], + device='cuda:1'), in_proj_covar=tensor([0.0309, 0.0329, 0.0295, 0.0328, 0.0329, 0.0283, 0.0450, 0.0308], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:1') +2023-02-09 04:19:42,469 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9092, 1.5420, 1.7613, 1.5055, 0.9926, 1.6105, 1.7769, 1.6862], + device='cuda:1'), covar=tensor([0.0569, 0.1239, 0.1637, 0.1434, 0.0611, 0.1404, 0.0675, 0.0639], + device='cuda:1'), in_proj_covar=tensor([0.0101, 0.0154, 0.0191, 0.0162, 0.0102, 0.0164, 0.0113, 0.0148], + device='cuda:1'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:1') +2023-02-09 04:20:03,783 INFO [train.py:901] (1/4) Epoch 30, batch 6500, loss[loss=0.1761, simple_loss=0.2696, pruned_loss=0.04124, over 8542.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2801, pruned_loss=0.0553, over 1616143.10 frames. ], batch size: 50, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:20:17,923 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.512e+02 2.582e+02 3.161e+02 3.840e+02 1.025e+03, threshold=6.322e+02, percent-clipped=7.0 +2023-02-09 04:20:30,131 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3523, 2.6213, 2.7488, 1.8254, 3.0781, 1.9824, 1.5923, 2.3095], + device='cuda:1'), covar=tensor([0.0994, 0.0467, 0.0367, 0.0892, 0.0551, 0.1007, 0.1097, 0.0606], + device='cuda:1'), in_proj_covar=tensor([0.0485, 0.0418, 0.0378, 0.0469, 0.0404, 0.0561, 0.0409, 0.0448], + device='cuda:1'), out_proj_covar=tensor([1.2810e-04, 1.0780e-04, 9.8317e-05, 1.2237e-04, 1.0558e-04, 1.5580e-04, + 1.0912e-04, 1.1716e-04], device='cuda:1') +2023-02-09 04:20:36,285 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=240950.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:20:38,865 INFO [train.py:901] (1/4) Epoch 30, batch 6550, loss[loss=0.1709, simple_loss=0.2553, pruned_loss=0.04322, over 7667.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2802, pruned_loss=0.05557, over 1614778.64 frames. ], batch size: 19, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:21:00,572 WARNING [train.py:1067] (1/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-02-09 04:21:01,494 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7730, 2.0513, 2.0673, 1.2901, 2.2054, 1.6607, 0.5468, 1.9924], + device='cuda:1'), covar=tensor([0.0725, 0.0427, 0.0389, 0.0726, 0.0470, 0.0971, 0.1052, 0.0330], + device='cuda:1'), in_proj_covar=tensor([0.0485, 0.0419, 0.0377, 0.0469, 0.0405, 0.0561, 0.0409, 0.0449], + device='cuda:1'), out_proj_covar=tensor([1.2821e-04, 1.0787e-04, 9.8151e-05, 1.2251e-04, 1.0565e-04, 1.5593e-04, + 1.0907e-04, 1.1719e-04], device='cuda:1') +2023-02-09 04:21:13,967 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.8493, 5.9230, 5.1801, 2.4982, 5.2345, 5.6587, 5.3041, 5.3440], + device='cuda:1'), covar=tensor([0.0488, 0.0335, 0.0763, 0.4243, 0.0747, 0.0695, 0.1062, 0.0469], + device='cuda:1'), in_proj_covar=tensor([0.0560, 0.0467, 0.0460, 0.0572, 0.0454, 0.0482, 0.0457, 0.0420], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:21:15,890 INFO [train.py:901] (1/4) Epoch 30, batch 6600, loss[loss=0.216, simple_loss=0.2998, pruned_loss=0.06615, over 8260.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2806, pruned_loss=0.05571, over 1616927.53 frames. ], batch size: 24, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:21:16,033 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241004.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:21:20,061 WARNING [train.py:1067] (1/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-02-09 04:21:28,329 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.35 vs. limit=2.0 +2023-02-09 04:21:29,318 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.0216, 2.2276, 1.8338, 2.9153, 1.3656, 1.7722, 2.1484, 2.2212], + device='cuda:1'), covar=tensor([0.0730, 0.0811, 0.0876, 0.0314, 0.1170, 0.1234, 0.0850, 0.0791], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0194, 0.0244, 0.0213, 0.0202, 0.0246, 0.0248, 0.0204], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 04:21:29,837 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.473e+02 2.233e+02 3.026e+02 3.930e+02 1.368e+03, threshold=6.053e+02, percent-clipped=4.0 +2023-02-09 04:21:51,501 INFO [train.py:901] (1/4) Epoch 30, batch 6650, loss[loss=0.1609, simple_loss=0.2382, pruned_loss=0.04182, over 7445.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.28, pruned_loss=0.05507, over 1618041.48 frames. ], batch size: 17, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:22:04,818 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=241073.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:22:23,560 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=241098.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:22:27,625 INFO [train.py:901] (1/4) Epoch 30, batch 6700, loss[loss=0.1915, simple_loss=0.284, pruned_loss=0.04947, over 7967.00 frames. ], tot_loss[loss=0.1947, simple_loss=0.2792, pruned_loss=0.05508, over 1619433.33 frames. ], batch size: 21, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:22:42,111 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.506e+02 2.246e+02 2.834e+02 3.422e+02 9.903e+02, threshold=5.667e+02, percent-clipped=4.0 +2023-02-09 04:23:04,036 INFO [train.py:901] (1/4) Epoch 30, batch 6750, loss[loss=0.2044, simple_loss=0.2941, pruned_loss=0.05734, over 8340.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2803, pruned_loss=0.05572, over 1625099.05 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:23:08,862 INFO [scaling.py:679] (1/4) Whitening: num_groups=1, num_channels=256, metric=4.47 vs. limit=5.0 +2023-02-09 04:23:39,180 INFO [train.py:901] (1/4) Epoch 30, batch 6800, loss[loss=0.1821, simple_loss=0.2664, pruned_loss=0.04891, over 6861.00 frames. ], tot_loss[loss=0.1943, simple_loss=0.2784, pruned_loss=0.05515, over 1615186.85 frames. ], batch size: 15, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:23:42,688 WARNING [train.py:1067] (1/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-02-09 04:23:53,781 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.574e+02 2.178e+02 2.682e+02 3.526e+02 7.087e+02, threshold=5.364e+02, percent-clipped=2.0 +2023-02-09 04:24:00,918 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3955, 2.0931, 1.6495, 2.0411, 1.7790, 1.4440, 1.6849, 1.6756], + device='cuda:1'), covar=tensor([0.1271, 0.0512, 0.1297, 0.0537, 0.0789, 0.1668, 0.0991, 0.0910], + device='cuda:1'), in_proj_covar=tensor([0.0362, 0.0246, 0.0347, 0.0316, 0.0304, 0.0350, 0.0350, 0.0322], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 04:24:15,452 INFO [train.py:901] (1/4) Epoch 30, batch 6850, loss[loss=0.2004, simple_loss=0.2776, pruned_loss=0.06165, over 8495.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2797, pruned_loss=0.05599, over 1613810.37 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:24:34,759 WARNING [train.py:1067] (1/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-02-09 04:24:43,872 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241294.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:24:50,575 INFO [train.py:901] (1/4) Epoch 30, batch 6900, loss[loss=0.18, simple_loss=0.2712, pruned_loss=0.04445, over 7976.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2801, pruned_loss=0.05646, over 1614191.55 frames. ], batch size: 21, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:25:05,717 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.702e+02 2.527e+02 3.094e+02 3.969e+02 8.004e+02, threshold=6.188e+02, percent-clipped=9.0 +2023-02-09 04:25:22,104 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241348.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:25:26,913 INFO [train.py:901] (1/4) Epoch 30, batch 6950, loss[loss=0.1788, simple_loss=0.2768, pruned_loss=0.04045, over 8330.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2789, pruned_loss=0.05555, over 1612684.85 frames. ], batch size: 25, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:25:46,735 WARNING [train.py:1067] (1/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-02-09 04:26:04,048 INFO [train.py:901] (1/4) Epoch 30, batch 7000, loss[loss=0.2564, simple_loss=0.3211, pruned_loss=0.09586, over 6845.00 frames. ], tot_loss[loss=0.1944, simple_loss=0.278, pruned_loss=0.05533, over 1604061.51 frames. ], batch size: 72, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:26:07,754 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=241409.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:26:17,959 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.728e+02 2.441e+02 2.932e+02 3.651e+02 7.920e+02, threshold=5.865e+02, percent-clipped=3.0 +2023-02-09 04:26:18,113 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241424.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:26:40,283 INFO [train.py:901] (1/4) Epoch 30, batch 7050, loss[loss=0.2227, simple_loss=0.3037, pruned_loss=0.07085, over 8295.00 frames. ], tot_loss[loss=0.1941, simple_loss=0.2781, pruned_loss=0.055, over 1606279.86 frames. ], batch size: 23, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:26:46,794 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=241463.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:27:16,741 INFO [train.py:901] (1/4) Epoch 30, batch 7100, loss[loss=0.1814, simple_loss=0.2796, pruned_loss=0.04159, over 8784.00 frames. ], tot_loss[loss=0.1942, simple_loss=0.2785, pruned_loss=0.05491, over 1609471.26 frames. ], batch size: 30, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:27:27,514 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6208, 2.0340, 3.1831, 1.4859, 2.4828, 2.0563, 1.6912, 2.5217], + device='cuda:1'), covar=tensor([0.2075, 0.2754, 0.0927, 0.4883, 0.1918, 0.3450, 0.2639, 0.2217], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0648, 0.0566, 0.0675, 0.0670, 0.0620, 0.0573, 0.0649], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:27:30,224 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6972, 1.8217, 1.6211, 2.2193, 1.0846, 1.4940, 1.7650, 1.8793], + device='cuda:1'), covar=tensor([0.0777, 0.0691, 0.0861, 0.0456, 0.1024, 0.1226, 0.0669, 0.0677], + device='cuda:1'), in_proj_covar=tensor([0.0231, 0.0193, 0.0244, 0.0213, 0.0202, 0.0245, 0.0247, 0.0203], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 04:27:30,723 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.720e+02 2.381e+02 2.857e+02 3.660e+02 8.579e+02, threshold=5.714e+02, percent-clipped=3.0 +2023-02-09 04:27:51,597 INFO [train.py:901] (1/4) Epoch 30, batch 7150, loss[loss=0.1522, simple_loss=0.2436, pruned_loss=0.03041, over 8083.00 frames. ], tot_loss[loss=0.1947, simple_loss=0.2791, pruned_loss=0.05517, over 1612502.23 frames. ], batch size: 21, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:28:00,173 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.5210, 2.3042, 1.8527, 2.2553, 1.9975, 1.5986, 1.8769, 1.9007], + device='cuda:1'), covar=tensor([0.1395, 0.0465, 0.1208, 0.0567, 0.0795, 0.1584, 0.0997, 0.0966], + device='cuda:1'), in_proj_covar=tensor([0.0362, 0.0246, 0.0348, 0.0317, 0.0305, 0.0351, 0.0352, 0.0323], + device='cuda:1'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:1') +2023-02-09 04:28:28,702 INFO [train.py:901] (1/4) Epoch 30, batch 7200, loss[loss=0.2028, simple_loss=0.2896, pruned_loss=0.05803, over 7822.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2796, pruned_loss=0.0554, over 1614361.25 frames. ], batch size: 20, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:28:43,487 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.677e+02 2.259e+02 2.765e+02 3.853e+02 1.030e+03, threshold=5.530e+02, percent-clipped=3.0 +2023-02-09 04:28:53,983 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241639.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:29:03,963 INFO [train.py:901] (1/4) Epoch 30, batch 7250, loss[loss=0.2041, simple_loss=0.2909, pruned_loss=0.0587, over 8080.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2795, pruned_loss=0.05546, over 1615135.23 frames. ], batch size: 21, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:29:11,970 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=241665.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:29:26,162 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7270, 1.6586, 2.1353, 1.3172, 1.3340, 2.0943, 0.3057, 1.2471], + device='cuda:1'), covar=tensor([0.1417, 0.0976, 0.0356, 0.1026, 0.2079, 0.0410, 0.1596, 0.1145], + device='cuda:1'), in_proj_covar=tensor([0.0203, 0.0206, 0.0138, 0.0223, 0.0277, 0.0149, 0.0174, 0.0200], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 04:29:30,389 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=241690.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:29:40,038 INFO [train.py:901] (1/4) Epoch 30, batch 7300, loss[loss=0.2376, simple_loss=0.324, pruned_loss=0.07564, over 8644.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2792, pruned_loss=0.05557, over 1611703.65 frames. ], batch size: 34, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:29:50,471 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=241719.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:29:53,736 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.652e+02 2.349e+02 2.997e+02 3.899e+02 6.597e+02, threshold=5.994e+02, percent-clipped=5.0 +2023-02-09 04:30:08,885 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=241744.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:30:15,572 INFO [train.py:901] (1/4) Epoch 30, batch 7350, loss[loss=0.2148, simple_loss=0.2954, pruned_loss=0.06708, over 8436.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2804, pruned_loss=0.05631, over 1614046.04 frames. ], batch size: 27, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:30:25,406 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241768.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:30:27,486 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([4.6287, 4.6607, 4.1177, 2.1871, 4.0695, 4.2207, 4.0801, 4.0761], + device='cuda:1'), covar=tensor([0.0632, 0.0456, 0.0971, 0.4289, 0.0932, 0.0886, 0.1165, 0.0748], + device='cuda:1'), in_proj_covar=tensor([0.0558, 0.0465, 0.0459, 0.0568, 0.0452, 0.0481, 0.0454, 0.0419], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:30:39,746 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-02-09 04:30:51,563 INFO [train.py:901] (1/4) Epoch 30, batch 7400, loss[loss=0.1798, simple_loss=0.2597, pruned_loss=0.04997, over 7545.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2802, pruned_loss=0.05607, over 1612180.17 frames. ], batch size: 18, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:30:59,776 WARNING [train.py:1067] (1/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-02-09 04:31:04,080 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241821.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:31:05,936 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.776e+02 2.497e+02 3.037e+02 3.880e+02 5.984e+02, threshold=6.074e+02, percent-clipped=0.0 +2023-02-09 04:31:24,324 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241849.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:31:27,626 INFO [train.py:901] (1/4) Epoch 30, batch 7450, loss[loss=0.1816, simple_loss=0.2659, pruned_loss=0.04864, over 7557.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2802, pruned_loss=0.05608, over 1610487.30 frames. ], batch size: 18, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:31:40,058 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-02-09 04:31:40,202 WARNING [train.py:1067] (1/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-02-09 04:31:48,158 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=241883.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:31:55,069 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.3951, 1.4843, 1.3942, 1.8365, 0.6602, 1.2781, 1.3194, 1.4871], + device='cuda:1'), covar=tensor([0.0927, 0.0758, 0.0973, 0.0489, 0.1183, 0.1424, 0.0802, 0.0741], + device='cuda:1'), in_proj_covar=tensor([0.0233, 0.0195, 0.0245, 0.0215, 0.0203, 0.0248, 0.0250, 0.0204], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 04:32:02,789 INFO [train.py:901] (1/4) Epoch 30, batch 7500, loss[loss=0.1585, simple_loss=0.2343, pruned_loss=0.04136, over 7213.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2808, pruned_loss=0.0564, over 1612789.71 frames. ], batch size: 16, lr: 2.49e-03, grad_scale: 16.0 +2023-02-09 04:32:18,961 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.524e+02 2.491e+02 2.852e+02 3.531e+02 9.058e+02, threshold=5.704e+02, percent-clipped=2.0 +2023-02-09 04:32:39,984 INFO [train.py:901] (1/4) Epoch 30, batch 7550, loss[loss=0.209, simple_loss=0.2906, pruned_loss=0.06368, over 8526.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2803, pruned_loss=0.05576, over 1615080.01 frames. ], batch size: 28, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:33:01,505 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241983.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:33:17,213 INFO [train.py:901] (1/4) Epoch 30, batch 7600, loss[loss=0.198, simple_loss=0.2873, pruned_loss=0.05436, over 8569.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2803, pruned_loss=0.05604, over 1610770.37 frames. ], batch size: 39, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:33:32,887 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.614e+02 2.414e+02 3.070e+02 3.745e+02 6.631e+02, threshold=6.140e+02, percent-clipped=3.0 +2023-02-09 04:33:45,095 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.7358, 2.1518, 3.2086, 1.5649, 2.5382, 2.2152, 1.8694, 2.5831], + device='cuda:1'), covar=tensor([0.1928, 0.2798, 0.0958, 0.4876, 0.1942, 0.3415, 0.2555, 0.2402], + device='cuda:1'), in_proj_covar=tensor([0.0545, 0.0649, 0.0566, 0.0678, 0.0670, 0.0619, 0.0573, 0.0648], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:33:54,462 INFO [train.py:901] (1/4) Epoch 30, batch 7650, loss[loss=0.219, simple_loss=0.2968, pruned_loss=0.07058, over 8604.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2818, pruned_loss=0.05715, over 1613109.57 frames. ], batch size: 31, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:34:25,714 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=242098.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:34:29,810 INFO [train.py:901] (1/4) Epoch 30, batch 7700, loss[loss=0.2164, simple_loss=0.305, pruned_loss=0.06391, over 8453.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2815, pruned_loss=0.05709, over 1615123.51 frames. ], batch size: 27, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:34:44,311 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.689e+02 2.431e+02 3.028e+02 3.722e+02 6.918e+02, threshold=6.057e+02, percent-clipped=1.0 +2023-02-09 04:34:46,822 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9706, 1.9936, 2.0131, 1.7179, 2.0576, 1.7498, 1.3681, 1.9653], + device='cuda:1'), covar=tensor([0.0423, 0.0360, 0.0258, 0.0420, 0.0387, 0.0565, 0.0677, 0.0248], + device='cuda:1'), in_proj_covar=tensor([0.0484, 0.0420, 0.0377, 0.0469, 0.0406, 0.0562, 0.0409, 0.0448], + device='cuda:1'), out_proj_covar=tensor([1.2781e-04, 1.0832e-04, 9.8124e-05, 1.2240e-04, 1.0608e-04, 1.5618e-04, + 1.0897e-04, 1.1705e-04], device='cuda:1') +2023-02-09 04:34:54,344 WARNING [train.py:1067] (1/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-02-09 04:34:55,280 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=242139.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:05,987 INFO [train.py:901] (1/4) Epoch 30, batch 7750, loss[loss=0.153, simple_loss=0.2353, pruned_loss=0.03536, over 7233.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2813, pruned_loss=0.05727, over 1613363.92 frames. ], batch size: 16, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:35:11,039 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.9729, 2.2865, 1.8917, 2.9406, 1.3246, 1.6330, 2.3040, 2.2071], + device='cuda:1'), covar=tensor([0.0844, 0.0766, 0.0859, 0.0366, 0.1186, 0.1343, 0.0784, 0.0781], + device='cuda:1'), in_proj_covar=tensor([0.0233, 0.0195, 0.0246, 0.0215, 0.0204, 0.0248, 0.0250, 0.0205], + device='cuda:1'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:1') +2023-02-09 04:35:11,679 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([5.5535, 5.6480, 4.9784, 2.5744, 4.9948, 5.2478, 5.2570, 5.1497], + device='cuda:1'), covar=tensor([0.0532, 0.0360, 0.0834, 0.4079, 0.0773, 0.0924, 0.0966, 0.0604], + device='cuda:1'), in_proj_covar=tensor([0.0557, 0.0463, 0.0456, 0.0567, 0.0451, 0.0480, 0.0453, 0.0418], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:1') +2023-02-09 04:35:13,189 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=242164.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:13,776 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=242165.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:18,598 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=242171.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:34,909 INFO [zipformer.py:1185] (1/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=242193.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:35:42,638 INFO [train.py:901] (1/4) Epoch 30, batch 7800, loss[loss=0.1666, simple_loss=0.2483, pruned_loss=0.04249, over 7806.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2819, pruned_loss=0.05727, over 1611061.37 frames. ], batch size: 19, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:35:57,874 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.684e+02 2.396e+02 3.014e+02 3.960e+02 8.063e+02, threshold=6.029e+02, percent-clipped=4.0 +2023-02-09 04:36:11,178 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=242244.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:36:18,002 INFO [train.py:901] (1/4) Epoch 30, batch 7850, loss[loss=0.2056, simple_loss=0.2894, pruned_loss=0.06095, over 8186.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2818, pruned_loss=0.05698, over 1614829.71 frames. ], batch size: 23, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:36:36,019 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=242280.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:36:52,298 INFO [train.py:901] (1/4) Epoch 30, batch 7900, loss[loss=0.1835, simple_loss=0.2607, pruned_loss=0.05319, over 7647.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2801, pruned_loss=0.05635, over 1610977.19 frames. ], batch size: 19, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:36:54,993 INFO [zipformer.py:1185] (1/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=242308.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:37:01,360 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-09 04:37:06,479 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.670e+02 2.359e+02 2.894e+02 3.889e+02 1.272e+03, threshold=5.788e+02, percent-clipped=10.0 +2023-02-09 04:37:08,036 INFO [zipformer.py:1185] (1/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=242327.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:37:11,720 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-02-09 04:37:14,231 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([2.3238, 2.1122, 2.6116, 2.2435, 2.6375, 2.4190, 2.1981, 1.5410], + device='cuda:1'), covar=tensor([0.6092, 0.5499, 0.2406, 0.4294, 0.2758, 0.3430, 0.2057, 0.5842], + device='cuda:1'), in_proj_covar=tensor([0.0974, 0.1040, 0.0857, 0.1017, 0.1038, 0.0951, 0.0783, 0.0864], + device='cuda:1'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:1') +2023-02-09 04:37:26,302 INFO [train.py:901] (1/4) Epoch 30, batch 7950, loss[loss=0.2019, simple_loss=0.2958, pruned_loss=0.05397, over 8318.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2807, pruned_loss=0.05708, over 1609311.64 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:37:26,517 INFO [zipformer.py:1185] (1/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=242354.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:37:37,023 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-02-09 04:37:42,302 INFO [zipformer.py:2431] (1/4) attn_weights_entropy = tensor([1.6532, 1.5411, 2.1142, 1.3838, 1.2176, 2.0756, 0.3489, 1.3051], + device='cuda:1'), covar=tensor([0.1458, 0.1177, 0.0392, 0.0964, 0.2321, 0.0424, 0.1856, 0.1164], + device='cuda:1'), in_proj_covar=tensor([0.0204, 0.0207, 0.0138, 0.0224, 0.0278, 0.0149, 0.0175, 0.0200], + device='cuda:1'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:1') +2023-02-09 04:37:42,514 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-02-09 04:37:43,556 INFO [zipformer.py:1185] (1/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=242379.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:38:00,602 INFO [train.py:901] (1/4) Epoch 30, batch 8000, loss[loss=0.2404, simple_loss=0.3118, pruned_loss=0.08451, over 8344.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2816, pruned_loss=0.05714, over 1613794.03 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:38:13,197 INFO [scaling.py:679] (1/4) Whitening: num_groups=8, num_channels=192, metric=1.57 vs. limit=2.0 +2023-02-09 04:38:14,883 INFO [optim.py:369] (1/4) Clipping_scale=2.0, grad-norm quartiles 1.710e+02 2.478e+02 2.968e+02 3.707e+02 1.083e+03, threshold=5.936e+02, percent-clipped=5.0 +2023-02-09 04:38:35,235 INFO [train.py:901] (1/4) Epoch 30, batch 8050, loss[loss=0.1985, simple_loss=0.2865, pruned_loss=0.05526, over 7938.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2806, pruned_loss=0.05733, over 1595420.49 frames. ], batch size: 20, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:38:58,134 INFO [train.py:1165] (1/4) Done!